In this study, we provide a high-resolution view of genome-wide transcriptional regulation in S. cerevisiae
. By coupling the NRO assay to high-throughput sequencing, we assayed the synthesis of nascent RNA in yeast under normal growth conditions and in response to heat stress. We measured RNA abundance and synthesis in parallel, which revealed that cellular RNA abundance in yeast under both growth conditions is predominantly controlled by the rate of RNA synthesis. In accord with previous degradation experiments, we observed distinct functional gene classes between stable and unstable transcripts (Friedel et al. 2009
; Grigull et al. 2004
; Narsai et al. 2007
; Wang et al. 2002
These data allowed us to examine the spatial distribution of RNA synthesis along transcripts, which showed that nascent transcription in yeast is enriched near TSSs, as it is in humans. The data further suggest that postinitiation regulation in yeast may contribute to the control of gene expression (Churchman and Weissman 2011
). Although recruitment of RNA polymerase to promoters appears to be rate-limiting for the large majority of yeast genes, transcription of up to 20% of yeast genes may be controlled at least in part at postinitiation steps. The mechanisms that determine polymerase pausing, arrest, or termination are not yet well understood. Many factors that modulate transcriptional pausing and the rate of elongation have been identified (Sims et al. 2004
). Some of them are restricted to a subset of eukaryotes, with no homologs identified in yeast. However, in a recent study, Rodriguez-Gil et al. (2010)
identified a subset of elongation-related factors (e.g.
DSIF, Mediator, and the RNA polymerase II subunit Rpb9
) that might influence the probability of RNA polymerase II pausing or arrest in yeast. In addition, the elongation factor TFIIS
) stimulates intrinsic RNA polymerase RNA cleavage activity and thus promotes elongation of arrested RNA polymerase in vivo
(Churchman and Weissman 2011
). Signals in the DNA or RNA may affect RNA elongation; for example, both a hairpin structure in the nascent RNA and a run of uridines at the 3′ end of a transcript interfere with elongation and promote transcription termination (Bengal and Aloni 1989
; Keene et al. 1999
; Palangat et al. 1998
The authors of recent studies have identified numerous cryptic unstable transcripts in yeast (Neil et al. 2009
; Xu et al. 2009
). Most of these transcripts were suggested to be byproducts of bidirectional transcription that originated from the same promoter but from distinct pre-initiation complexes (Neil et al. 2009
). We observed little bidirectional transcription in either the NRO or total RNA samples. However, detection of cryptic unstable transcripts relied on the use of a strain defective in the nuclear exosome machinery (Neil et al. 2009
; Xu et al. 2009
). Thus, it is possible that most of the antisense transcription was undetectable in the strain used in our study, which is competent for RNA turnover. For nondivergent promoters, with an average intergenic distance of approximately 342 bp (Dujon 1996
), a weak antisense signal arises at approximately −200 bp from the TSS, much upstream of the core promoter. Similarly, for divergent promoters, the antisense signal does not begin to accumulate until approximately −150 bp from the TSS. The distances at which antisense transcription is originated suggest that it likely results from the activity of a separate transcription pre-initiation complex assembled at a different core promoter.
An important biological question is how cells respond and cope with rapid changes in their environment, such as exposure to elevated temperatures (heat shock). Several genome-wide studies revealed that this response involves an increase in the abundance of transcripts from genes that are collectively termed “heat shock-inducible” (Gasch et al. 2000
; Hahn et al. 2004
). Earlier studies of the Drosophila
hsp70 gene found that posttranscriptional regulation that affects RNA stability plays a role in the observed increase of hsp70 RNA levels in response to heat shock (Theodorakis and Morimoto 1987
). Other studies of heat shock-inducible genes in Drosophila
and humans found rapid transcriptional activation after heat shock, suggesting that this activation also contributes to increased RNA levels of heat shock-inducible genes (Mathur et al. 1994
; O’Brien and Lis 1993
). However, in previous studies investigators have not assessed global changes in nascent transcription in response to heat shock. Here, we found that both transcriptional activation and RNA stabilization likely play a role in the increased abundance of RNAs after heat shock treatment. Most of the heat shock-inducible genes (~74%) showed both an increase in nascent transcription and an increase in RNA abundance. Only approximately 14% of these genes increased their RNA abundance without increasing nascent transcription, indicative of RNA stabilization as the mechanism.
In summary, we used the NRO assay to provide a high-resolution view of transcriptional regulation in yeast. We applied this assay to examine the correlation between synthesis and abundance of transcripts, the distribution of polymerase activity along transcripts, and the genome-wide changes in yeast transcription in response to heat shock treatment. Further investigations may use similar approaches to characterize other modes of yeast transcriptional control under different growth conditions and in response to other environmental perturbations.