Nucleotide excision repair hotspots and coldspots of UV-induced DNA damage in the human genome

Abstract

1We recently developed high-throughput sequencing approaches, eXcision Repair 2 sequencing (XR-seq) and Damage-seq, to generate genome-wide mapping of DNA 3 excision repair and damage formation, respectively, with single-nucleotide resolution. 4Here, we used time-course XR-seq data to profile UV-induced excision repair dynamics, 5 paired with Damage-seq data to quantify the overall induced DNA damage. We identified 6 genome-wide repair hotspots exhibiting high-level nucleotide excision repair immediately 7 after UV irradiation. We show that such repair hotspots do not result from hypersensitivity 8 to DNA damage, and are thus not damage hotspots. We find that the earliest repair occurs 9 preferentially in promoters and enhancers from open-chromatin regions. The repair 10 hotspots are also significantly enriched for frequently interacting regions and super-11 enhancers, both of which are themselves hotspots for local chromatin interactions. 12

Dissecting regulatory pathways for transcription recovery following DNA damage reveals a non-canonical function of the histone chaperone HIRA

“…5a), which is a major determinant of the UV damage load. Indeed, UV lesions are distributed throughout the human genome with very few detectable damage hotspots [77][78][79] , thus the UV damage load is generally proportional to gene length and UV dose. Considering that 10 J/m 2 UVC irradiation generates 1 UV lesion/6 kb on double-stranded DNA 80 , the expected damage load on the transcribed strand, where UV lesions block RNAPII progression, is around 1 UV lesion/12 kb.…”

Dissecting regulatory pathways for transcription recovery following DNA damage reveals a non-canonical function of the histone chaperone HIRA

“…5a), which is a major determinant of the UV damage load. Indeed, UV lesions are distributed throughout the human genome with very few detectable damage hotspots [76][77][78] , thus the UV damage load is generally proportional to gene length and UV dose. Considering that 10 J/m 2 UVC irradiation generates 1 UV lesion/6 kb on double-stranded DNA 79 , the expected damage load on the transcribed strand, where UV lesions block RNAPII progression, is around 1 UV lesion/12 kb.…”

The major mechanism of melanoma mutations is based on deamination of cytosine in pyrimidine dimers as determined by circle damage sequencing

“…Since deamination of CPDs is strictly a chemical reaction, substantial levels of deaminated CPDs were found at the 0-hour time point because the deamination reaction is expected to continue during DNA processing in vitro. The reasons for the lower levels of deaminated CPDs at the TSS could be the reduced propensity of TSS sequences to be transiently single stranded (which favors cytosine deamination), the shielding of these regions by general transcription factors, and/or the particularly rapid repair of these regions (28,(36)(37)(38), which may proceed to some extent during the 48-hour deamination time span even at the relatively high UV dose that we used. At the TES, there is a spike of total CPDs but a dip of deaminated CPDs, which is likely based on DNA sequence features, i.e., AT richness of these regions (Fig.…”