Ultraviolet light can damage DNA, triggering a general shutdown of gene transcription — yet some genes are activated by UV light. An investigation of this counter-intuitive behaviour reveals a surprising gene-regulation mechanism.
DNA damage can lead to mutations, cancer and even cell death, and activates several biological processes1,2. These include: DNA repair that removes lesions from the double helix; checkpoints that arrest cell-cycle progression to prevent the transmission of damaged DNA to daughter cells; apoptosis, a form of cell death that eliminates cells with heavily damaged genomes; and a transcription response that changes the cellular RNA profile. Much is known about how cells orchestrate these processes, but the transcription response is arguably the least well understood of the four. Writing in Cell, Williamson et al.3 describe a mechanism for the transcription response to DNA damage caused by ultraviolet light. Their findings reveal a remarkable circuit that triggers the formation of non-coding RNA molecules from a gene. Moreover, these molecules oppose the action of the protein that is produced from the same gene in the absence of UV damage.
DNA-damaging agents induce the transcription of specific gene classes. Several genes can be induced by more than one DNA-damaging agent, whereas others are induced mainly by one agent. The resulting gene products are associated with many different cellular processes, including DNA repair, intercellular signalling and responses to oxidative stress.
Irradiation by UV light elicits a response from a large subset of damage-induced genes, through a series of intracellular processes that starts at or near the cell membrane4. But a general shutdown of transcription also occurs soon after UV irradiation. This is because UV exposure causes chemical modifications in DNA known as pyrimidine dimers (PDs). When present on the transcribed strand of a gene, PDs stall the enzyme RNA polymerase II (RNAPII), which elongates nascent RNA chains. A process called transcription-coupled nucleotide excision repair rapidly removes these defects, allowing fast resumption of transcription. Cells that fail to resume transcription are eliminated by the ultimate DNA damage-response pathway: apoptosis5,6.
How can some genes be activated by UV light if transcription in general has been shut down by PDs? It is tempting to speculate that PDs on the transcribed strand of UV-activated genes are repaired particularly quickly, or that fewer PDs form in those genes. The number of PDs that form in a gene is usually proportional to gene size, because PDs form almost randomly in typical DNA sequences. It is therefore interesting that the genes activated by cytotoxic doses of UV light are compact, and contain a small number of short introns7 (non-coding sequences).
Williamson et al. now reveal a regulatory mechanism for UV-activated genes. They report that the transcription of several such genes by RNAPII proceeds slowly after UV exposure, and is restricted to regions of DNA that are close to promoters (promoter sequences are those that initiate gene transcription). Intriguingly, these changes in RNAPII behaviour correlate with the emergence of alternative splicing — that is, with a change in the way that nascent RNA is processed. This triggers the formation of short, non-coding transcripts, rather than the formation of longer, protein-encoding messenger RNAs (Fig. 1).
The authors focused their efforts on the ASCC3 gene, which researchers from the same group had previously implicated in the UV-damage response8. The ASCC3 protein encoded by the long mRNA transcribed from this gene suppresses transcription in the late stages of the response. Unexpectedly, the authors found that the short ASCC3 transcript produced by alternative splicing is a regulatory non-coding RNA that is required for the recovery of transcription.
The mechanisms by which the ASCC3 protein regulates transcription are poorly understood. However, Williamson et al. provide some clues by showing that ASCC3 interacts with both RNAPII and a transcription-repair coupling factor. The researchers have thus uncovered a regulatory circuit of the UV response, in which one RNA transcript produces a protein, whereas the other, shorter transcript resulting from alternative splicing opposes that protein's activity.
Future studies of the function of short RNA molecules promise to be fascinating, because our current understanding of such RNAs as activators of transcription is in its infancy. Williamson et al. say they are now screening for genes that encode factors required for transcription shutdown after DNA damage. It will be interesting to see how these factors modulate transcription — is there a connection with transcription-coupled repair, or do they belong to an as-yet-unknown pathway? And what is their function in cell survival?
The authors report that almost 80% of the transcripts that undergo processing in response to UV light can be accounted for by a simple model in which alternative splicing occurs unusually close to the promoter. But the remaining 20% or so include sequences from farther along the gene, and are not easily explained by a model based solely on slow versus fast RNAPII rates.
Another challenge will therefore be to determine the mechanisms underlying the formation of this minor population of mRNAs. Factors beyond simple RNAPII speed that might affect the choice of splice sites include compaction of the gene (chromatin structure) and competition between the folding of newly transcribed RNA and the binding of proteins that control splice-site choice9. In the meantime, Williamson et al. have added a new level of understanding to the mechanisms used by mammalian cells to regulate gene expression in response to UV irradiation. The discovery of RNA-processing circuitry that responds to cellular stress opens up the possibility that analogous gene-regulatory pathways have roles in the response to other cellular cues.Footnote 1
Sancar, A., Lindsey-Boltz, L. A., Ünsal-Kaçmaz, K. & Linn, S. Annu. Rev. Biochem. 73, 39–85 (2004).
Hanawalt, P. C. DNA Repair 36, 2–7 (2015).
Williamson, L. Cell 168, 843–855 (2017).
Herrlich, P., Ponta, H. & Rahmsdorf, H. J. Rev. Physiol. Biochem. Pharmacol. 119, 187–223 (1992).
Spivak, G. DNA Repair 36, 13–18 (2015).
Mullenders, L. DNA Repair 36, 28–35 (2015).
McKay, B. C. et al. Proc. Natl Acad. Sci. USA 101, 6582–6586 (2004).
Boeing, S. et al. Cell Rep. 15, 1597–1610 (2016).
Saldi, T., Cortazar, M. A., Sheridan, R. M. & Bentley, D. L. J. Mol. Biol. 428, 2623–2635 (2016).
Related links in Nature Research
About this article
Breast Cancer (2018)
Chemical Communications (2017)