lncRNA-ZFAS1, an emerging gate-keeper in DNA damage-dependent transcriptional regulation

doi:10.21203/rs.3.rs-3233912/v1

Download PDF

Article

lncRNA-ZFAS1, an emerging gate-keeper in DNA damage-dependent transcriptional regulation

https://doi.org/10.21203/rs.3.rs-3233912/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Numerous long noncoding RNAs (lncRNAs) are generated in response to external stimuli, but the scope and functions of such activity are not known. Here, we provide insight into how the transcription of lncRNAs are connected to DNA damage response by identifying a lncRNA ZFAS1, which is required for cell cycle arrest, transcription regulation and DNA repair. Mechanistically, ZFAS1 facilitates the changing hyper-phosphorylated RNAPII (hyper) isoforms around transcription initiation sites by directly targeting the regulated genes. We revealed extensive transcription shutdown and concomitant stimulated engagement of RNAPII-Ser2P are crucial for repair and cell survival upon genotoxic stress. Finally, ZFAS1knockout in mice dampened transcription-coupled nucleotide excision repair (TC-NER) and led to kidney dysplasia. Our study extends the understanding of lncRNAs in DNA damage repair (DDR) and implies a protective role of lncRNA against DDR-deficient developmental disorders.

Biological sciences/Molecular biology/DNA damage and repair/Nucleotide excision repair

Biological sciences/Molecular biology/Non-coding RNAs/Long non-coding RNAs

The hereditary information encoded in DNA sequence is intrinsically susceptible to alterations, being continually threatened by a variety of genotoxic perturbations(Jackson and Bartek, 2009). To safeguard the stability of the genome, eukaryotic cells have evolved a set of sophisticated surveillance system that controls several aspects of the cellular response, including the detection of DNA lesions, a temporary cell cycle arrest, regulation of transcription, and the repair of the damaged DNA(Helleday et al., 2014; Jackson and Bartek, 2009). To date, numerous DNA repair mechanisms that operate in different stages of the cell cycle and under distinctive types of DNA lesions have been well characterized(Giglia-Mari et al., 2011). When cells encounter stressful conditions, the DNA damage response (DDR) precisely coordinates the multiple cellular events by activating two parallel routes: (1) the DNA damage checkpoint, a cell cycle blockage that restrains chromosome segregation to avoid inappropriate transmission of the genetic information(Bartek and Lukas, 2007; Chao et al., 2017; Sancar et al., 2004) and (2) the stalling and degradation of transcribing RNA polymerase II (RNAPII), a DNA damage senser that triggers DNA repair(Gregersen and Svejstrup, 2018; Lavigne et al., 2017). On the other hand, after DNA damage repair, timely checkpoint recovery and restoration of transcription is crucial for continued cell proliferation(Gregersen and Svejstrup, 2018). However, it is still poorly understood how the DNA damage checkpoints and stalled RNAPII molecules convert a very limited amount of molecular-level information (even a single DNA lesion) in the context of an otherwise genome into regulation that halts and resumes the cell-cycle engine in a coordinated way.

Eukaryotic cells preferentially remove bulky DNA lesions such as those generated by UV irradiation from the transcribed strand of active genes by transcription-coupled nucleotide excision repair (TCR)(Gaillard and Aguilera, 2013; Hanawalt and Spivak, 2008), while the rest of the DNA lesions are eliminated by a slower repair pathway global genome repair (GGR) without strand specificity(Mellon et al., 1987). NER (nucleotide excision repair) plays an important role during G1 phase, if left unrepaired, bulky DNA lesions can block DNA polymerases. Studies in rodents reveal that G2 arrest is manifest under conditions of proficient TCR only in combination with deficient GGR, indicating that TCR enables replication of damaged DNA and elicits progression of cells through the S phase, leading to G2 arrest resulting from persisting DNA damage load in the genome requiring proficient GGR(van Oosten et al., 2000). As the focal point of TCR, the bulk of RNAPII effectively shields the DNA lesions from recognition by DNA repair factors(Brueckner et al., 2007) and degrades in a last-resort pathway when TCR fails(Wilson et al., 2013). Beyond these well-studied functions, tremendous progress has recently been made in the elucidation of the mechanistic intricacies of DNA damage-stalled RNAPII. The rationale for the genome-wide shutdown of transcription as a consequence of the physical blocking of RNAPII progress by DNA lesions has been challenged by the uninterrupted detection of functionally essential stress-dependent boost in the 5’ nascent RNA activity globally(Andrade-Lima et al., 2015; Lavigne et al., 2017). Mapping of chromatin state upon DNA damage illustrates a model of constant recruitment of new RNAPII molecules to guarantee a pool of non-engaged polymerases, enabling prolonged and efficient DNA lesion scanning of the whole transcribed genome(Liakos et al., 2020). This rationale is not mutually exclusive with the functions of RNAPII ubiquitination and degradation. Although several proteins have been linked to RNAPII ubiquitination in yeast and human cells(Anindya et al., 2007), the precise role of human RNAPII ubiquitination in addition to acting as a last-resort pathway to remove and degrade RNAPII from DNA damage sites has only been explored most recently. Actually, site-specific ubiquitination of RNAPII can promote recruitment of TFIIH, mediate transcription shutdown, and initiate TCR(Nakazawa et al., 2020; Vidakovic et al., 2020). Together, these findings substantiate the impact of transcription on the spatio-temporal orchestration of DNA repair activities, making it imperative to explore the mechanisms in detail to unveil the causative connections between transcription, DNA damage checkpoint, and DNA repair.

The human genome transcribes numerous long noncoding RNAs (lncRNAs), and only a very few of them have been functionally characterized in certain processes and diseases(Johnsson et al., 2014; Rinn and Chang, 2012; Wapinski and Chang, 2011). It has been suggested that a large fraction of lncRNAs represent byproducts of promiscuous transcription rather than independently functional RNAs(Struhl, 2007). For example, DNA damage-induced transcription of non-coding regulatory loci is proposed to keep their transcription factor-binding sequences in check and guarantee accurate regulation of topologically associated mRNA genes(Liakos et al., 2020; Sabarinathan et al., 2016; Scruggs et al., 2015). To resolve this issue, detailed functional studies are required to establish the molecular activity and biological role of noncoding RNAs of interest. Although evidence for function of ncRNAs as a group is lacking, several lncRNAs have been implicated being able to affect the cellular response to DNA damage. The prime examples are in the genomic loci of cell cycle genes. Two lncRNAs (named PANDA and DINO) transcribed antisense to CDKN1A are induced by DNA damage and regulate the activity of distantly located genes in trans. PANDA negatively regulates apoptosis by blocking the transcription factor NF-YA(Hung et al., 2011), while DINO can amplify DNA damage signaling by mediating the stabilization of p53(Schmitt et al., 2016). Another example is the ASCC3 gene which can express both coding and non-coding transcript isoforms. The two isoforms act in opposition to balance the recovery of transcription after UV-induced DNA damage(Williamson et al., 2017). Inspired by these examples, we hypothesized that more stress-dependent lncRNAs may serve as component of the DNA damage response to be functionally essentially involved in the orchestration of DNA repair that have yet to be discovered.

In this study, we reveal a map of extensive lncRNA transcription during DDR by using synchronized cells, leading to the unexpected identification of a poorly characterized mammalian lncRNA-ZFAS1. We describe that ZFAS1 functions as a key player of cellular response to DNA damage in both human and rodent cells by fine tuning RNAPII kinetics, suggesting a lncRNA-dependent transcriptional regulatory axis that maintains genomic stability upon DNA damage in mammalian cells.

DNA damage checkpoints activated by severe stress can trigger permanent cellular outcomes such as apoptosis or senescence(Nyberg et al., 2002), meanwhile, failure to balance between cell cycle arrest and recovery can cause genome instability and impaired cell proliferation after DDR(Sogo et al., 2002). It has long been known that cells in different cell-cycle phases show drastic differences in survival under the same levels of ultraviolet light irradiation, which induces DNA adducts(Han et al., 1971). Although the molecular pathways of the DDR in each cell-cycle phase employ redundant mechanisms(Shaltiel et al., 2015), it has been proposed that each phase elicits distinct DNA checkpoint dynamics that link to the predicted cell-cycle outcomes(Chao et al., 2017). To explore the mechanisms by which cells regulate their G1/S phase transcriptional network to execute the cell cycle arrest and DNA damage checkpoint response, we exposed synchronized human fibroblasts at the G1/S boundary to 10 J/m² UV-C, which were then subjected to RNA sequencing at the indicated time periods for recovery. Upon UV treatment, the synchronized cells released from the G1/S block exhibited a prolongation of the S phase and an extended late S-phase-G2 arrest, corresponding to the checkpoint commitment time at the G1/S transition and S/G2 transition (Fig. 1A). This delayed cell cycle progression was precisely coordinated with the dynamic expression of the rescheduled Cyclin A and Cyclin B (Fig. 1B). Furthermore, the substantial removal of CPD (cyclobutane pyrimidine dimers) lesions from UV-exposed synchronized cells was accomplished by 24 h after release from the G1/S block (Figure S1A). The detailed kinetic analysis of nascent RNA transcript synthesis post UV-C irradiation revealed that transcription inhibition and recovery were best achieved at 3 h and 24 h respectively (Figure S1B). We therefore initiated the study in an attempt to identify human lncRNAs that regulate the DNA damage response by using this well-established cellular system.

Orchestration of gene expression dynamics during DNA damage response

To systematically illustrate a map of extensive transcription programs that may orchestrate the DNA damage-induced cell cycle checkpoint dynamics, we performed lncRNA-seq to obtain the expression profiles of lncRNAs and mRNAs in UV-irradiated G1/S-phase-synchronized cells. Approximately 63,370 human protein-coding transcripts and 23,227 lncRNAs (including annotated and novel lncRNAs) were identified in this study, with 5,044 lncRNAs under the criteria of FPKM ≥ 0.2 in at least one sample (Figure S1C and Table S1). HC analysis (unsupervised hierarchical clustering) and PCA (principal component analysis) of lncRNAs as compared to protein-coding mRNAs revealed significant differences in RNA-seq profiles (Figs. 1C and S1D). In particular, the discernable boundaries (or phases) that may demarcate the key stages with distinct gene expression changes during DDR were observed at the mRNA expression levels (with PC1 reflecting the treatment types, PC2 and PC3 reflecting the progression stage) (see left panel in Figure S1D). Furthermore, the irradiated cells can be partitioned into two distinguished phases: phase 1 (UV_3 h to UV_12 h) and phase 2 (UV_24 h to UV_96 h) (see left panel in Fig. 1C). In contrast, lncRNAs showed less pronounced stage-specific distribution of RNA-seq reads across the cell cycle (see right panel in Fig. 1C and right panel in Figure S1D).

Firstly, we focused on the role of transcriptional activation during cell cycle progression in the absence of genotoxic stress. We detected protein-coding genes as significantly induced at some point in time after release from the G1/S phase (henceforth referred to as temporal genes) and grouped them into six manually defined clusters using hierarchical clustering (Figure S2A). To identify representative genes in particular stages and understand their biological significance, we performed gene set enrichment analysis of the resulting clusters of temporal profiles. Genes that are immediately activated after release from the G1/S block (3 h; FDR < 0.05) are enriched for nucleosome assembly, indicating an association with DNA replication during S phase. Genes induced at the early time point (6 h; FDR < 0.05) are enriched for cell-cell adhesion and microtubule organization, marking the interphase-mitosis (G2/M) transition of the cell cycle progression (Figure S2A and Table S2). This result implies that cell-cycle commitment coincides with genome-wide changes in transcription. Inspection of the well-defined stage-specific markers, such as HIST1H1D and TBP (required for chromatin organization involved in S phase), and AURKA and CLIP2 (responsible for microtubule formation occurring during G2/M transition), confirmed their predicted cell cycle phase-specific expression (Figure S2B). Secondly, we attempted to examine the transcriptional change in response to UV-C irradiation by applying the same algorithm. In total, we identified 8,421 stage-specific transcripts which can be partitioned into seven groups based on their expression patterns across the DDR process (Figure S2C and Table S3). The genes that were immediately upregulated upon UV irradiation were mainly associated with cellular response to stress (3 h after UV; FDR < 0.05). The late phase (72 h after UV; FDR < 0.05) contains genes involved in cell morphogenesis and cell growth (Figure S2C and Table S4). Overall, these data revealed a highly coordinated transcriptional program in human fibroblast cells during DDR. We subsequently surveyed the expression changes of three important protein families during the DNA damage-induced cell cycle progression. As expected, we observed cyclin genes, CCNA1 (G1/S phase regulator)(Ji et al., 2005) and CCNB1 (G2/M transition controller)(Müllers et al., 2014), to be upregulated at 6 h and 9 h after release from G1/S phase, respectively. UV-C exposure significantly delayed this activation until 24 h after release (Figure S3A). Mitosis marker AURKA (responsible for spindle assembly)(Goldenson and Crispino, 2015) and its kinase PLK1(Joukov and De Nicolo, 2018) were upregulated at 9 h in the absence of DNA damage, with postponed induction after UV-C treatment (Figure S3B). DNA damage-dependent transcriptional repressor KLF10 was immediately induced upon UV-C irradiation and transactivated its target CDKN1A to stimulate temporal confinement of cell cycle progression in response to DNA damage (Figure S3C). These results further confirmed that changes in expression of key genes related to the specific cell cycle stages during the DDR can be well recapitulated in our system.

Extensive and regulated lncRNA transcription of human fibroblasts during DNA damage-induced cell cycle progression

Although a systematic identification of lncRNAs that might be involved in DNA damage-induced cell-cycle progression has never been reported, potential roles for individual lncRNAs in DDR or cell-cycle regulation have been demonstrated before. In our data, we detected a few previously studied cell cycle phase-specific lncRNAs, such as LUCAT1 and MANCR(Yildirim et al., 2020), to display the highest expression levels during S phase (3–6 h after release in the non-irradiated cells), with significantly delayed induction upon UV exposure (24–48 h after release in the irradiated cells) (Figure S4A). The well-defined DNA damage response lncRNAs, TP53TG1(Diaz-Lagares et al., 2016) and NEAT1(Adriaens et al., 2016), were immediately activated upon UV-C irradiation, showing undisturbed expression patterns during normal cell cycle progression without genotoxic stress (Figure S4B). These results attested the reliability of our data to identify lncRNAs during cell cycle progression and DDR. Given that lncRNAs with much shorter half-lives are less stable than protein-coding mRNAs (Clark et al., 2012), we reasoned that lncRNAs which are meant to function at a narrow time window are likely to be activated at that specific stage. We detected 7,464 temporal lncRNAs which can be partitioned into six clusters based on their expression patterns in the absence of genotoxic stress (Fig. 2A and Table S5). We further explored the genes and pathways that these stage-specific lncRNAs may regulate by examining the biological functions of mRNAs co-expressed or co-localized with these lncRNAs. This analysis revealed that the expression levels for lncRNAs grouped into cluster 3 and cluster 6 had a general correspondence with the concurrent stage-specific events and included fitting GO biological term enrichment for cell cycle regulation (lncRNAs in cluster 3 are associated with cytokinesis, lncRNAs in cluster 6 are associated with DNA replication; Fig. 2A and Table S5). Upon UV-C irradiation, 6,500 lncRNAs were identified to be upregulated at one specific time point, among which lncRNAs immediately activated at 3 h upon UV-C exposure were predominately involved in the DNA damage response pathways, while genes co-localized with the lncRNAs that were highly expressed at 12 h after UV-C treatment were enriched for functional groups in cell cycle regulation (Fig. 2B and Table S6). High correlation of the dynamic expression patterns of these lncRNAs with different cellular conditions and distinct cell cycle phases suggest that these lncRNAs may be functional in the DNA damage associated cell cycle regulation.

To gain further insight into the nature of these temporal lncRNAs during DNA damage-induced cell cycle progression, we assessed their molecular features including exon number and gene length. In comparison to the background exon number of all lncRNAs, we did not observe any consistent over- or under-represented differences across all samples (Figure S4C). In contrast, the stage-specific lncRNAs had a distinct gene length, with the early (clusters 2–4, 3–9 h upon UV-C irradiation) and late (clusters 7–9, 48–96 h after UV-C exposure) DNA damage-induced lncRNAs generally being shorter than other lncRNAs activated during the intermediate phase (clusters 5–6, 12–24 h in response to UV-C treatment) (Figures S4D-E).

To investigate the functional relevance of these DNA damage-induced lncRNAs, we manually surveyed those stage-specific lncRNAs and selected three of them (SNHG15, SLC25A34-AS1, and ZFAS1) based on their abundance and dynamic expression profiles. All of them showed RT-PCR-validated periodic expression during the cell cycle, attesting to our dataset analysis (Figure S5A). Further examination of their nearby protein-coding genes yielded ZFAS1 as a particularly interesting DNA damage-induced lncRNA (Figs. 3A and S5B), in that it was immediately activated in response to genotoxic stress and exhibited highest expression level by 24 h after UV-C irradiation (corresponding to the cell cycle recovery after DNA damage repair). More importantly, contrary to ZFAS1, its anti-sense protein-coding gene ZNFX1 showed dramatic decrease in expression after UV-C exposure (Figures S5C and S5D), implying that the induction of ZFAS1 upon DNA damage does not derive from overlapping transcription of other active elements in the antisense direction.

ZFAS1 is a conserved, DNA damage-induced lncRNA

ZFAS1 is located approximately 583 base pairs upstream of the ZNFX1 TSS, and is modestly conserved across many mammalian species, underscoring its potential functional importance (Figure S6A). Rapid amplification of 5’ and 3’ complementary DNA ends (RACE) revealed two transcript variants of ZFAS1 [528 and 466 nucleotides (nt)] (Figures S6B-C), of which the 466-nt short variant was much more highly expressed with or without genotoxic stress (Fig. 3B). Indeed, RNA-seq data from the NONCODE project(Zhao et al., 2021) also showed highly expression of ZFAS1 short variant in various human tissues (Figure S6D). Like most other intergenic lncRNAs, ZFAS1 exhibited very low protein coding potential (Figure S7A). To further confirm that ZFAS1 is a genuine lncRNA without coding capacity, a putative ORF in ZFAS1 was predicted by ORF finder (http://www.ncbi.nlm.nih.gov/orffinder/), and the Western blot analysis revealed that neither the full length CDS nor the putative ORF of ZFAS1 had coding ability (Figs. 3C and S7B). The cellular distribution of ZFAS1 was then evaluated by FISH and confocal microscopy at the indicated recovery time after UV-C exposure. ZFAS1 in the unsynchronized MRC5_VA cells was found to be located both in the nucleus and cytoplasmic compartment in a DNA damage independent manner (Fig. 3D), which was further confirmed quantitatively by RNA-seq of chromatin/cytoplasm cellular fractionation (Figure S7C)(Liu et al., 2022). ZFAS1 was also induced upon cisplatin damage (forming DNA adducts as UV damage does and also triggering NER), albeit at lower levels (Figures S7D and S7E). In contrast, treatment with anti-tumor agents (5-FU or Doxorubicin) did not have effects on the expression of ZFAS1 (Figures S7D and S7E).

Given the activation of ZFAS1 was specific and fairly stable upon UV-C exposure, whether this was the result of epigenetic changes that occurred during DNA damage response was further investigated. ChIP and qPCR were performed to examine the major histone modifications and RNAPII occupancy in undisturbed and UV-irradiated MRC5_VA cells. Increased occupancy of RNAPII around the ZFAS1 transcription start site (TSS) after UV-C exposure was observed (Fig. 3E). Two histone modifications H3K4me3 and H3K27Ac that positively regulate transcription were also enhanced within the promoter-proximal region of ZFAS1 during DNA damage response (Fig. 3E). Accordingly, results from our previous studies(Liu et al., 2021) revealed an increase in chromatin accessibility as measured by ATAC-seq near the TSS of ZFAS1 in response to UV-C treatment (Figure S8A). Next, genome-wide mapping of 5^mC methylation levels at the ZFAS1 locus was conducted to investigate the correlation between DNA methylation changes and the expression dynamics of ZFAS1(Liu et al., 2021). Notably, compared to the CpG enriched regulatory regions like enhancers and promoters flanking the 5’ end of ZFAS1, which exhibited constant 5^mC levels during DNA damage response (data not shown), the UV-irradiated cells showed a more overt decrease in 5^mC methylation at the 3’ end of intragenic regions of ZFAS1 (Figure S8B). It has been shown in recent years that DNA N6-methyladenine modification (6mA) as a new epigenetic mark is present in human genome(Xiao et al., 2018). Therefore, 6mA-IP-qPCR technology was used to examine the DNA 6mA modification of ZFAS1 in response to genotoxic stress. The results showed that ZFAS1 harbored more obvious enrichment of 6mA modification compared with the negative control actin gene (Fig. 3F). Intriguingly, UV-C exposure significantly increased the 6mA modification levels of ZFAS1 (Fig. 3F), suggesting a potential role of 6mA in actively modulating gene expression of non-coding RNAs. On the basis of these results, we proposed that the accessible chromatin structure and active histone modifications on ZFAS1 loci contribute to the stimulated expression upon UV-C exposure.

Knockdown of ZFAS1 impairs TC-NER

To investigate the role of ZFAS1 in cellular response to DNA damage, lentivirus-mediated RNA interference (RNAi) was used to knockdown ZFAS1 in human fibroblast cells (Figure S9A). Cells depleted of ZFAS1 were hypersensitive to UV-C irradiation (Fig. 4A). Accordingly, ZFAS1 overexpression resulted in higher level of cell viability in response to DNA damage as compared to the wild type controls (Figures S9B and S9C). Genome-wide measurements of repair kinetics revealed that UV-induced (6 − 4) photoproducts are repaired quickly, while cyclobutane pyrimidine dimers (CPDs) generally require 12–48 h for complete removal by NER(Adar et al., 2016), the latter are more suitable to be measured to access TC-NER ability of the cell. Our result showed that ZFAS1 depleted MRC5_VA cells failed to accomplish CPD repair within 24 hours upon UV-C irradiation (Fig. 4B). On the contrary, MRC5_VA cells overexpressing ZFAS1 were equally proficient in removal of CPD lesions as compared to the wild-type cells (Figure S9D).

The γH2A.X foci (phosphorylation of the histone variant H2A.X on serine 139) are formed presumably at sites of DSBs (DNA double strand breaks)(Mah et al., 2010). Although UV-C exposure does not directly induce DSBs, it has been reported that γH2A.X is triggered by UV-induced DNA repair intermediates in all phases of the cell cycle(Hanasoge and Ljungman, 2007) and GG-NER pathway (not TC-NER) plays a major role in H2A.X phosphorylation(O'Driscoll et al., 2003). Thus, the kinetics of γH2A.X was widely used to monitor the replication and transcription stress in cells. In line with previous reports, γH2A.X was immediately detected as early as 30 min, continued to increase through 3 h and diminished at 24 h post-irradiation in the wild type cells (Fig. 4C). There was no discernable difference in the levels of H2A.X phosphorylation at early time points after UV-C treatment between the wild-type and ZFAS1-depleted MRC5_VA cells, though a high level of constitutive expression of γH2A.X was still observed in ZFAS1 depleted MRC5_VA cells at 24 h after UV-C treatment (Fig. 4C). Consistent with this observation, the persistent high levels of DNA breaks in ZFAS1 depleted MRC5_VA cells post-irradiation were also confirmed using an alkaline comet assay (Figure S9E).

An important role of the DNA-damage checkpoints is to sense DNA damage and mediate temporary cell-cycle arrest to allow time to repair DNA lesions. To test the requirement for ZFAS1 in cell cycle arrest in response to DNA damage, asynchronous wild-type or ZFAS1 depleted MRC5_VA cells were blocked at the G1/S border, and then exposed by UV-C irradiation prior to the release into the S-phase. Flow cytometry analysis showed that wild type cells exhibited an abrupt and sustained pause of the cell cycle during S phase within the first 9 hours before transition to the next phase (Fig. 4D, see the large population of cells remaining at S phase from 3 to 9 hrs after release for wild type cells upon UV irradiation). Strikingly, knockdown of ZFAS1 resulted in markedly decreased retention of cells during S phase and increased number of cells traversing into G2 phase (Fig. 4D, see a group of cells entering G2 phase 3–6 h after release for ZFAS1 depleted cells), indicating that ZFAS1 might be required for cell cycle arrest in response to DNA damage. It is important to remark that ZFAS1 might be indirectly involved in the correct of DNA lesions via its ability to positively regulate the DNA damage checkpoint. Notably, a DNA damage independent cell cycle progression failure was also detected in cells after depletion of ZFAS1 (Figure S9F, see the delayed G2/M transition for the ZFAS1 depleted MRC5_VA cells).

In order to investigate whether the GG-NER capacity of ZFAS1 depleted cells was intact, the unscheduled DNA synthesis (UDS) assay was performed. A non-radioactive technique by incorporating 5-Ethynyl-2’-deoxyuridine (EdU) to measure UDS has been shown to be sufficient for routine XP (Xeroderma Pigmentosa, a disease mainly characterized by GG-NER defects) screening(Limsirichaikul et al., 2009). The cell cycle independent UDS resulting from repair synthesis can be easily differentiated from replicative DNA synthesis in S-phase cells by distinct nuclear fluorescent intensities (Figure S9G, see the nuclei with weak fluorescent intensity indicated by arrows). This result showed that both wild-type and ZFAS1 depleted cells exhibited indistinguishable UDS activity after UV-C irradiation (Figure S9G), revealing that ZFAS1-deficient cells are only compromised in the TCR pathway of NER.

ZFAS1 functions by modulating global transcription shutdown upon DNA damage

Cells carrying mutations in some DNA repair factor genes are unable to recover RNA synthesis from damaged genes and unleash attenuated RNAPII transcription elongation after UV treatment, while wild-type cells become transcriptionally active only a few hours upon UV irradiation(Gregersen and Svejstrup, 2018; Mayne and Lehmann, 1982). Since the signature of ZFAS1 depleted cells was reminiscent of DNA repair factor deficiency in response to DNA damage, it is reasonable to predict that like many other DNA repair factors such as XPF and CSB(Le May et al., 2010; Wang et al., 2014), ZFAS1 might also be a player involved in the global transcriptional response to UV irradiation. For measuring global transcription, cells were labeled with 5-ethynyl-uridine (EU) 3 h and 24 h after UV-C irradiation for 30 minutes (Figure S10A). The staining of nascent transcripts with EU revealed that transcription across the genome was essentially shut down 3 h after UV-C irradiation in both wild-type and ZFAS1-depleted cells. Unexpectedly, transcription levels were tellingly fully restored 24 h after UV-C exposure in both cell lines (Figure S10A). To further characterize the role of ZFAS1 in transcription regulation after DNA damage, the nascent RNA synthesis was examined at early time intervals (15 min to 3 hr after UV-C exposure) to gain insight in the early stages of transcriptional shutdown and restart. Global transcription was immediately repressed by 15 mins after UV-C treatment in the wild-type cells and decreased further by 1 hr after UV-C exposure (Fig. 5A). On the contrary and strikingly, knockdown of ZFAS1 resulted in high global transcription levels at early time-points after UV-C irradiation, as indicated by a reduced percentage of lowly transcribing cells and an increase in EU incorporation signified by an obvious shift of the histogram to the right within 1 hr in response to UV-C exposure as compared to the wild-type cells (see the blue curve in the histogram in Fig. 5A). This unexpected and inspiring role of ZFAS1 as an entity to keep transcription repressed after DNA damage was further confirmed by qRT-PCR in combination with DRB treatment. Cells were first treated with the transcription elongation inhibitor DRB to reversibly block newly initiated RNAPII, and then UV-irradiated, followed by inhibitor removal to concomitantly release synchronized RNAPII, so that the nascent transcription at different intron-exon junctions can be assessed. In line with those observed before(Saponaro et al., 2014), in the untreated wild-type cells, RNAPII progressed to the first promoter-proximal exon-intron junction of the human KIFAP3 gene 20 min after DRB removal and to an exon-intron junction 153 kilobases downstream from the TSS after 2 hr (Figure S10B). Upon UV-C irradiation, a dramatic reduction in RNAPII progress was observed, with nascent RNA moving very slightly further forward to the first exon-intron junction after 50 min and to the very last exon-intron junction after more than 24 hr (Fig. 5B). Tellingly, the arrested pre-mRNA levels at both the first exon-intron junction and the last exon-intron junction restored markedly earlier in cells depleted for ZFAS1 after DNA damage (Fig. 5B, 20 min for RNAPII to reach Exon1-Intron1, 22 hr for RNAPII to reach Exon19-Intron19 in cells lacking ZFAS1). Importantly and surprisingly, the effect of ZFAS1 on global transcription in the absence of exogenous genotoxic stress was the opposite of that observed upon DNA damage (Figure S10B, comparing purple to red graphs, blue to green graphs), in agreement with the effect of ZFAS1 lacking on cell cycle progression (Figure S9F). These results revealed the dual roles of ZFAS1 in activating and repressing gene expression in a manner that depends on the presence of DNA damage.

ZFAS1 regulates the distribution of Ser2P- and Ser5P-modified RNAPII upon UV-C

We were initially puzzled by the finding that the halted RNA synthesis was recovered more rapidly in cells lacking ZFAS1 after UV irradiation, yet the ZFAS1 deficiency resulted in the failure of DNA damage repair. To obtain a precise and genome-wide view into the molecular events controlling transcription reorganization in response to genotoxic stress, ChIP-seq was performed in the wild-type and ZFAS1 depleted MRC5_VA cells using antibodies against CTD phosphorylation-specific RPB1. Without UV-C irradiation, the distribution of RNAPII-Ser5P was comparable between the wild type and ZFAS1 depleted cells (Figures S10C and S10D), indicating that the effect that ZFAS1 deficiency slows down general transcription as signified in DRB treatment assay (Figure S10B) does not lie under a decrease of pre-initiating RNAPIIo levels. At 3 hr after UV-C exposure, in line with the previously observed global loss of RNAPII at TSSs (Liakos et al., 2020), there was an obvious decrease in RNAPII-Ser5P read densities at promoter-proximal region in the WT cells, which were reduced further down at 6 hr and started to shift back at 12 hr after UV irradiation (Fig. 5C, compare red, green and blue to black curves for WT cells). The ZFAS1-depleted cells displayed a similar RNAPII-Ser5P distribution pattern at all time points except 6 hr (Fig. 5C, compare red and green to black curves for ZFAS1-KD cells), at which further reduction of RNAPII-Ser5P occupancy was absent, implying a potential inefficient transcription arrest. On the other hand, an increased concomitant enrichment of RNAPII-Ser2P near the 3’ of TSS was observed in the WT cells at 3 hr and 6 hr after UV stress (Fig. 5D, compare red and green to black curves for WT cells). These results mirror previous analyses(Nakazawa et al., 2020), but were in striking contrast to those obtained in cells when ZFAS1 was knocked down. Here, the RNAPII-Ser2P occupancy was unchanged at all time points in cells lacking ZFAS1 upon UV-C exposure (Fig. 5D). Taking into consideration recent evidence that supports the model of DNA damage-induced continuity of transcription initiation of short RNA transcripts(Andrade-Lima et al., 2015; Lavigne et al., 2017; Liakos et al., 2020), it is tempting to assume that the immediate loss of RNAPII-Ser5P and gain of RNAPII-Ser2P at TSSs (or rapid exchange of RNAPIIo isoforms) observed in the WT cells would help guarantee the uninterrupted global release of scanning enzymes into transcribable DNA sequences containing lesions. Thus, the defective transcriptional response at early time points upon DNA damage in cells lacking ZFAS1 might be critically linked to the DNA repair deficiency (Fig. 5E). The important exchange of RNAPIIo isoforms was further assessed by presenting the difference in occupancy profiles as log2 fold change in treated compared to untreated cells (Figs. 5F-G and Figures S10E-F). The wild-type cells displayed a substantial depletion of promoter-bound RNAPII-Ser5P and a concomitant increase of elongating RNAPII-Ser2P downstream of the TSS after UV irradiation (up panels in Figs. 5F-G). On the contrary, the reorganization of RNAPIIo isoforms was not observed in the ZFAS1-depleted cells (bottom panels in Figs. 5F-G). We also calculated the ratio of RNAPII-Ser5P read densities between the gene body and the promoter-proximal regions (termed as the travelling ratio or escape index)(Lavigne et al., 2017; Saponaro et al., 2014) before and after DNA damage. In accordance with the previous observations, a transient yet significant increase of promoter escape indexes at early time points during transcription recovery after UV stress (3 to 6 h post-irradiation) was detected only in the wild-type cells (Fig. 5H). Accordingly, the results of our genome-wide analysis were consistent with the evidence obtained by peak calling of RNAPII phosphorylation isoforms, showing that in the wild-type cells, RNAPII-Ser5P levels were reduced at 3 hr to 6 hr and restored at 12 hr (up panel in Fig. 5I, Table S7). Meanwhile, RNAPII-Ser2P levels were elevated drastically at early time points and returned to the basal level at 12 hr in response to UV-C exposure (up panel in Fig. 5J, Table S7). Reciprocally, there was no obvious loss of RNAPII-Ser5P or gain of RNAPII-Ser2P peaks in cells lacking ZFAS1 during the post-UV period (bottom panels in Figs. 5I-J).

Taken together, these results consolidate further the previous reported evidence(Liakos et al., 2020; Nakazawa et al., 2020) and reveal an unanticipated functional insight into a transcription-dependent defense mechanism modulated by ZFAS1: upon DNA damage, it is crucial to maintain low levels of RNAPII-Ser5P at the promoter-proximal regions to prevent genome instability threat, meanwhile firing elongating RNAPII-Ser2P to accelerate damage-sensing. In this scenario, ZFAS1 actively preserves genome integrity and promotes DNA lesion repair by adjusting the dynamically engaged RNAPIIo isoforms.

Increased genomic occupancy of ZFAS1 is associated with dynamic distribution of RNAPIIo upon DNA damage

Interestingly, although an impaired DNA damage-induced genome-wide rearrangement of RNAPIIo isoforms upon ZFAS1 depletion was observed, ZFAS1 knockdown had little effect on total mRNA levels after UV stress. First of all, ZFAS1 depletion did not change the expression patterns of its adjacent protein-coding gene ZNFX1 (Figure S11A). Principal component analysis (PCA) from RNA-seq revealed that all samples formed three distinct clusters irrespective of cell origins (Figure S11B), therein the separation was primarily based on treatment conditions. Furthermore, the ZFAS1-depleted cells displayed similar transcriptional responses to DNA damage in terms of the number of differentially expressed genes as compared to the WT cells (Figure S11C). We surmised that the recurring arrested RNAPII Ser5P at the promoter-proximal regions post-irradiation could be key to understanding ZFAS1 function, so we initially focused on the repressed genes in response to DNA damage. ATF3 was recently proved to target and regulate a large proportion of late response genes following UV irradiation(Epanchintsev et al., 2017), it was reasonable to predict that ZFAS1 may regulate the stability of ATF3 and, consequently, the expression of its targeted genes. RNA-seq analysis showed that more than 30% of downregulated genes at 3 hr after UV-C exposure in both the wild-type and ZFAS1-depleted cells were targeted by ATF3 (left panels in Figure S11D). Moreover, ChIP-seq analysis revealed that among the genes harboring reduced levels of RNAPII-Ser5P at 3 hr after UV irradiation in the two cell lines, more than 20% of which were targeted by ATF3 (right panels in Figure S11D). In accordance with these observations, ZFAS1 depletion had no effect on the activation and degradation of ATF3 during DDR (Figure S11E). These results suggest that ZFAS1 is not involved for ATF3-dependent transcriptional repression after DNA damage.

To determine the underlying molecular mechanisms by which ZFAS1 manipulates RNAPIIo isoforms in response to DNA damage, chromatin isolation by RNA purification (ChIRP) was applied to map ZFAS1’s chromatin occupancy before and 3 h after UV-C irradiation. ZFAS1 RNA was specifically retrieved with “’even” and “odd” pools of biotinylated capture probes along with associated DNA (Figures S12A-B). Notably, robust changes in ZFAS1 distribution on the genome following DNA damage were observed, with an increase of the ZFAS1 occupancy enriched over gene bodies (Figures S12C and S12D), implying a potential function of ZFAS1 that acts by modulating RNAPII traveling upon UV-C irradiation. By using the same pipeline as reported before(Chu et al., 2011), we identified 5,108 ZFAS1 raw peaks in the undisturbed cells, 7,319 raw peaks (+ UV) and 5,938 raw peaks (+ UV2) in the irradiated cells for the two biological replicates (Table S8). Furthermore, by using the stringent analysis pipeline to screen for “true peaks”, 665 ZFAS1 occupancy sites were identified in the control cells, comparing to 823 true peaks (874 true peaks from UV2 replicate) in the irradiated cells (Table S9). To compare ZFAS1 binding sites across samples, the read intensities of ctrl-even ZFAS1 true peaks were ranked from high to low, which is favorably correlated with the ZFAS1 occupancy sites enriched by ctrl-odd probes (panels I-II in Fig. 6A). Next, the level of ZFAS1 mark in the corresponding regions in the UV-irradiated cells were plotted side-by-side with that of ctrl-even (panels III-VI in Fig. 6A), which showed similar pattern, with the global level significantly higher (Figs. 6A and 6B). Accordingly, obvious increase of ZFAS1 raw peaks was also observed in cells after DNA damage (Figures S13A and S13B). ZFAS1 binding sites are enriched in genic regions, notably regions annotated as promoters and introns (Fig. 6C). Intriguingly, ZFAS1 prominently localizes at the TSSs (almost exclusively increased at the region of promoter < = 1 kb) in the UV-irradiated cells (Fig. 6C), strongly indicating a role for ZFAS1 in the regulation of transcription initiation/preinitiation during the post-UV period. Meanwhile, ZFAS1 binding events are focal, as typical ZFAS1 peaks are no more than a few hundred base pairs (Figures S13C and S13D), a pattern reminiscent of transcription factors. Compared to the genes bound by ZFAS1 in the undisturbed cells, the genes with ZFAS1 genomic occupancy after UV-C irradiation were significantly enriched for the GO terms related to dephosphorylation (Fig. 6D and Table S10). When the most stringent thresholds were applied, 111 common genes with bona fide peaks were identified in the both independent replicates upon UV-C irradiation (Fig. 6E). Among these genes, only five of them were found to be bound by ZFAS1 in control cells without UV stress (Fig. 6F). Furthermore, the binding sites on these so called “ZFAS1-gain” genes upon UV irradiation were equally located at promoter, intragenic and intergenic loci (Figure S13E). Intriguingly, genes associated with ZFAS1-gain loci exhibited aberrant spatiotemporal distributions of RNAPIIo isoforms in response to UV-C treatment in the ZFAS1-depleted cells as compared to the WT cells (Figs. 6G-H and Figures S14A-D). Analysis of UV-induced ZFAS1 binding sites revealed significantly enriched DNA motifs that are quite similar to the motifs of four transcription factors (KLF5, p53, Pax7, and RarA) (Figure S14E). LncRNAs can regulate gene expression both in cis and in trans. Colocalization analysis between ZFAS1 and these identified transcription factors may help decipher the way that ZFAS1 accesses the genome and mediates the RNAPIIo dynamics in response to UV-C irradiation.

Given the small number of ZFAS1 “true peaks” identified and the large number of genes with dynamic RNAPIIo occupancy, only a small proportion (~ 7%) of genes containing dynamic distributions of RNAPIIo were bound by ZFAS1 (red bars in Figure S14F). Inversely and strikingly, a large proportion of ZFAS1 target genes exhibited prominent increase of CTD-ser2P occupancy (> 58%) at early time points post-UV irradiation (blue bars in Figure S14F). More importantly, most of the ZFAS1-bound genes with spatiotemporal distribution changes of RNAPIIo failed to respond after ZFAS1 knockdown (white bars in Figure S14F). Last but not least, we attempted to verify the correlation between ZFAS1 binding events and DNA lesion repair. It has been found recently that the human fibroblast genome contains the hyper UV-sensitive regions, most of which reside within genes(Premi et al., 2019). The result showed no enrichment of CPD hyper-hotspots upon UV-C exposure in the ZFAS1-bound genes (Figure S14G). Taken together, the results suggest that UV-induced genome-wide ZFAS1 recruitment is not coupled to DNA damage recognition, rather serves as a chromatin bookmark that licenses transcription to reorganize upon DNA damage.

We next performed ChIRP assays with biotinylated ZFAS1, followed by mass spectrometry (ChIRP-MS) in MRC5_VA cells to search for potential ZFAS1-interactors. The results revealed that ZFAS1 can interact with multiple proteins in the absence or presence of UV-C exposure (Figure S15A). Gene ontology analysis illustrates high overall enrichment of nucleosome-related categories in the dataset (Figure S15B). SSRP1, NME and XRCC5 (the first two proteins are UV-dependent interactors) were initially selected as ZFAS1-associated proteins (Figure S15C), unfortunately none of them was confirmed by independent immunoblot (data not shown). RIP assay only verified the specificity of the interaction between NME1 and ZFAS1 (Figure S15D). The poorly understood signaling pathway containing NME1 and ZFAS1 in DDR merits further investigation. For instance, whether administration of NME1 inhibitors can impair DDR functionally in a manner similar to ZFAS1 knockdown.

Germline disruption of mouse Zfas1 impairs DDR

Mouse Zfas1 transcripts can be detected in seven major mouse tissues being examined (except pancreas, see Figure S16A). In some tissues (like liver and kidney), Zfas1 was highly expressed during the embryonic stage of mice. Whereas in other tissues (like brain and spleen), Zfas1 exhibited a strong expression signal in postnatal mice. We generated Zfas1 knockout (KO) C57BL/6 mice with CRISPR/Cas9 technology to characterize the function of the mouse Zfas1 in the DNA damage response (Fig. 7A). The targeting construct was designed to achieve two purposes: (i) effectively disrupt the expression of Zfas1 and (ii) avoid interference of endogenous anti-sense Znfx1 (Figure S16B). All the mice were born with expected Mendelian inheritance ratios, indicating that deficiency in Zfas1 does not interfere with embryonic or neonatal development of mice. Zfas1 KO mice develop normally and are indistinguishable from the heterozygous and wild-type littermates in terms of body weights (Figure S16C). Since in mice, Zfas1 is highly expressed in liver and kidney in embryonic stage (Figure S16A), we have carefully evaluated the development and functions of these tissues. Intriguingly, young (2-month-old) Zfas1 KO mice displayed small kidneys (Figure S16D). Histological analysis revealed that the Zfas1 KO mice kidneys contained a subset of glomeruli with lesions including lobulated phenomenon and proliferation of mesangial cells (Figure S16E). However, no data showed that the Zfas1 KO mice ultimately succumb to renal failure due to the significant glomerular damage observed at early developmental stage. In order to investigate the appearance of potential neurological aberrations in the Zfas1 KO mice, several behavioral tests were performed on 8-week-old mice. In the open-field test, the male Zfas1 KO mice demonstrated increases in total distance travelled and corresponding movement speed, indicative of hyperactivity (Figure S16F). The Morris water maze test was applied to measure the cognitive functioning of the Zfas1 KO mice. The learning ability and long-term spatial memory of the KO mice was not impaired (Figures S16G and S16H). No obvious morphological abnormalities were observed in the cerebrum and cerebellum of the 2-month-old ZFAS1 KO mice, with similar expression levels of GFAP (marker for astrocytes) in the cerebral cortex (Figure S16I). Collectively, these data indicate that neurologic dysfunction of the Zfas1 KO mice is present in very mild form.

Next, to confirm that targeted disruption of the mouse Zfas1, like human ZFAS1, resulted in defective transcription-coupled NER, a series of DNA repair parameters were examined in primary mouse lung fibroblasts. The cellular survival assay after exposure to increasing UV doses showed that the Zfas1^−/− mouse fibroblasts were UV-sensitive as compared to the WT controls (Fig. 7B). The mouse cells’ ability to repair CPD lesions induced by UV-C exposure was affected by Zfas1 depletion (Fig. 7B). This inefficient CPD repair is likely derived from deficiency in TC-NER pathway, as is evident from Figure S16J, the UDS levels observed in Zfas1^−/− fibroblasts were comparable to that in the wild-type mice, indicating that they are global genome repair-proficient. In line with the previous observation that human ZFAS1 depletion affects the kinetics of γH2A.X formation in response to DNA damage (Fig. 4C), a high level of DSBs was observed in the mouse Zfas1^−/− cells relative to those of the WT fibroblasts 24 h after UV-C exposure (Fig. 7D). Furthermore, DNA damage-triggered reduced RNA synthesis was sustained for at least 24 hrs in the WT cells prior to restoration. In the strikingly contrast, normal transcription levels were almost fully recovered within 24 h after UV-C exposure in the Zfas1^−/− fibroblasts (Fig. 7E). The photosensitivity of Zfas1^−/− mice was also tested by exposing the shaven dorsal skin of the wild-type and the homozygous mutant littermates to UV-B light at the dose of 500 J/m²/day for 4 consecutive days. Compared to the wild type mice, Zfas1^−/− mice developed severe UV burns within a week (Fig. 7F). Histological analysis of skin sections of Zfas1^−/− mice revealed epidermal hyperplasia, which was not observed in the UV-exposed skin of the WT mice (Fig. 7G). Taken together, these data revealed that, analogous to human ZFAS1 depleted cells, Zfas1-deficient mouse fibroblasts are impaired in TC-NER by interfering with transcription shutdown and recovery.

Although the genome-wide shutdown of transcription after UV-C irradiation in human cells has been known to occur for a considerable time(Mayne and Lehmann, 1982), the precise mechanisms and functions underlying this event remain poorly understood. In this report, we have identified the lncRNA ZFAS1 as an important component of the UV-induced DNA damage response, regulating the widespread UV-triggered processing and distribution of RNAPIIo on chromatin which is essential for both TC-NER and cell survival after genotoxic stress.

Transcription regulation after DNA damage: difference in lncRNAs versus mRNAs

A cell’s response to DNA damage depends on its exact cell-cycle position by utilizing an all-or-none checkpoint mechanism in G1 and G2 phase or a graded slowdown model in S phase, which in turn leads to heterogeneous cell fate outcomes(Chao et al., 2017). Here, we used synchronously proliferating human fibroblast cells that experience DNA damage at the same time during a given cell-cycle phase to capture the dynamic transcriptional reprogramming that occur in response to DNA damage, with particular emphasis on lncRNAs. Although the molecular mechanisms underlying major transcriptional events, including RNAP II stalling and degradation and transcriptional silencing and recovery, have recently been intensely implicated in DNA repair(Liakos et al., 2020; Nakazawa et al., 2020; Vidakovic et al., 2020), the impact of this coordinated transcription on DNA repair activities and cellular survival has rarely been studied. By employing a full time-course RNA-seq to cover the whole DNA damage response, we have been able to scrutinize potential differences between protein-coding and lncRNA genes. In general, protein-coding genes show higher ability to work in concert with the cell-cycle phase-specific checkpoint dynamics (Fig. 1A). Thus, activation of DNA damage response genes at early time points is a hallmark of recognition and removal of DNA damage, whereas induction of cell cycle genes at late time points precisely defines the resumption of cell cycle progression (Figures S2 and S3). In contrast, the stage-specific transcriptional dynamics of protein-coding genes that appear to be tightly regulated are generally absent for lncRNAs (Fig. 1A).

In our previous studies, we have revealed that distinct from protein-coding genes that transcripts with shorter length prefer to be recovered first, repression of lncRNA transcription after UV exposure is inactivated first on transcripts with longer length(Liu et al., 2022). In line with previous reports that virtually all genes induced by DNA damage are short(McKay et al., 2004), here we also observed that the bulk of lncRNAs induced by UV during the early and late stages across the DNA damage response are relatively shorter (Figure S4). All these findings unravel that lncRNAs might use a largely uncommon mechanism or a set of unique factors to induce or resume transcription following DNA damage.

The importance of proper transcription shutdown and recovery on TC-NER

The cellular TC-NER-compromised phenotype of ZFAS1-deficient cells was demonstrated by the sensitivity to UV-C irradiation, delayed or inefficient removal of UV-induced DNA lesions, and sustained expression of UV-triggered H2AX phosphorylation (Fig. 4). However, the impaired recovery of RNA synthesis, a general hallmark of TC-NER disorders(Mayne and Lehmann, 1982; van der Horst et al., 1997), is absent in the ZFAS1-deficient cells (Fig. 5A). It may appear counter-intuitive that ZFAS1 depletion results in both reduced cellular survival and faster transcription recovery in response to DNA damage. Strikingly, in ZFAS1-depleted cells, we found genome-wide preservation (slight but not significant increase) of RNAPII S5P levels at 3–6 h in response to UV at promoter-proximal regions. Reciprocally, there was no gain of RNAPII S2P immediately after UV exposure (Figs. 5C-E). We therefore proposed that the defective global transcription during the early steps of DDR in ZFAS1-depleted cells does not occur due to de novo recruitment of RNAPII-hypo molecules, but because of the arrest of the already promoter-bound elongating RNAPII. Accordingly, the ineffective transcription shutdown and aberrant cell cycle arrest fail to provide enough time to allow DNA repair, and might ultimately promote cells with damaged DNA to survive the cycle. Whether ZFAS1 regulates transcription shutdown by directly manipulating checkpoint factors requires further investigation.

Ultimate transcriptional recovery is crucial for cells to survive UV irradiation, a process that fails in individuals suffering from Cockayne syndrome(Mayne and Lehmann, 1982; Wang et al., 2014). Several factors have been implicated to play roles in re-establishing the active transcriptional state upon DNA damage(Adam et al., 2013; Dinant et al., 2013; Proietti-De-Santis et al., 2006; Williamson et al., 2017). Nonetheless, the importance of transcription shutdown in mammalian cells is not fully understood, because the precise events involved in RNAPII metabolism occurring within a few hours upon DNA damage have been multitude and controversial. While earlier studies revealed that new transcription initiation events are transiently inhibited after UV irradiation which coincides with the ensuing fast release of promoter-proximal RNAPII and global loss of the hypophosphorylated form of RNAPII(Epanchintsev et al., 2017; Gregersen and Svejstrup, 2018; Heine et al., 2008; Lans et al., 2019; Rockx et al., 2000). On the other hand, recent work found that new RNAPII molecules are constantly recruited to PICs post-UV in parallel with a gain in chromatin accessibility and conservation of active histone marks at the TSSs. This phenomenon is underlined by the identification of stress-dependent initiation-associated short RNA synthesis(Liakos et al., 2020; Liu et al., 2022). Our findings reveal the severely reduced RNAPII S5P at TSSs and the concomitant increase in RNAPII S2P downstream of TSS in WT cells (Fig. 5). We thus hypothesized that the apparent loss of RNAPII S5P may represent the disengagement of a readily phosphorylated RNAPII molecules from DNA template after damage recognition, while enhancement of RNAPII S2P stands for active transcription initiation that maximizes lesion scanning and removal. However, we do not exclude the possibility that increase of RNAPII S2P may not be due to newly recruitment of RNAPII, but rather due to stress-triggered modification change of RNAPII S5P which is already loaded at active genes upon UV irradiation. In the ZFAS1-deficient cells, the rapid exchange of RNAPII phosphorylation isoforms upon UV exposure was interrupted, leading to failure of DNA repair and cell death. Our data underscores the crucial role for correct transcription shutdown and supports a recent model that the fine tuning of the free RNAPII pool regulates the global transcriptional response to UV irradiation(Nakazawa et al., 2020; Vidakovic et al., 2020).

ZFAS1 -chromatin interactions guide DNA damage-induced RNAPII transcription

ChIRP-seq revealed that ZFAS1 occupancy sites in the genome are focal and numerous with increasing tendency toward the promoter-proximal regions, peaking at TSS sites in response to UV-C irradiation (Fig. 6C). More importantly, there appeared to be a reciprocal gain in binding sites following UV-C exposure (Figs. 6A and 6E). GO term analysis confirmed that these genes with enhanced ZFAS1 binding affinity are enriched in dephosphorylation process (Fig. 6D). Among the individual genes that are targeted by ZFAS1 upon DNA damage, it is worth noting to detect the tyrosine phosphatase EYA4 which specifically dephosphorylates Tyr-142 of histone H2AX(Cook et al., 2009) and the adaptor protein NCK1 which mediates the catalytic activity of protein phosphatase I (PP1)(Kremer et al., 2007). Most of the DNA damage checkpoint routes rely on the active phosphorylation of their components, recent discoveries have put emphasis on the important roles of protein phosphatases as active enzymes that are essential to re-establish the dephosphorylation status of multiple DDR targets during checkpoint arrest and cell cycle re-entry(Campos and Clemente-Blanco, 2020). In general, an involvement in the fine-tuning regulation of phosphorylation events might potentially help explain the aberrant cell cycle arrest occurring in the absence of ZFAS1 upon UV exposure.

Comparing to histone modifications which often broadly occupy certain genomic elements(Rando and Chang, 2009), the interspersed and gene-selective feature of ZFAS1 occupancy pattern more resembles transcription factors, which binds across gene bodies with focal peaks of high occupancy at TSS sites upon DNA damage. These results imply that ZFAS1 may access the genome in a highly discriminating fashion. However, the existence of the specific regulatory elements for ZFAS1 target sites in the genome is still obscure. We found that the ZFAS1 motif has similarities to several transcription factors binding motif but in a more degenerate pattern (Figure S14E). The ZFAS1 motif may serve as the binding site of a protein that recruits ZFAS1, or indirectly configure a chromatin state (DNA structure) that facilitates ZFAS1 binding. Additional studies are required to evaluate these hypotheses, which are now possible due to the identification of the candidate DNA binding protein NME1. By performing ChIRP-MS, we isolated NME1 assembled with ZFAS1 in a UV dependent manner (Figure S15C-D). The roles of NME1 in DNA proofreading and repair has been well documented due to its 3’-5’ exonuclease activity(Ma et al., 2004). NME1 has also been identified as potential transcription factor through its DNA-binding activities, although lacking a canonical nuclear localization signal(Choudhuri et al., 2010). Indeed, recent studies indicate that the presence of NME1 in the nucleus is induced by DNA-damaging agents(Jarrett et al., 2012). Here we suggested a new potential nuclear function of NME1 as a complex with ZFAS1 to mediate transcriptional program in response to UV exposure. Due to lacking of NME1 occupancy map after DNA damage, it is unclear whether the information for target gene selectivity resides in ZFAS1, which then recruits NME1 to chromatin or vice versa. Although the precise, molecular mechanism underlying the function of ZFAS1-NME1 complex is unknown, we uncovered persuasive correlation between DNA binding events of ZFAS1 and RNAPIIo reorganization in response to DNA damage.

ZFAS1 KO mice exhibit impaired DDR: implications for developmental disorders

We find that the knockout of mouse Zfas1 results in pronounced TC-NER deficiency and kidney abnormalities. Strikingly, unlike patients with DDR deficiency usually develop devastating features of impaired neurological functions(Madabhushi et al., 2014), conversely, the DDR mediated by Zfas1 does not exhibit neuronal deficits. In support of this notion, increasing evidence has revealed that animal models with mutations in key DNA repair genes often exhibit tissue-specific phenotypes. For example, in contrast to human Cockayne syndrome individuals, TC-NER-deficient Csa^−/− or Csb^−/− mice dot not show an obvious neurological phenotype(Jaarsma et al., 2013; van der Horst et al., 1997). A similar situation is observed in Fanconi anemia repair pathway-deficient mice, which do not develop FA(Bogliolo and Surralles, 2015; Wong and Buchwald, 2002). Until now, we know surprisingly little about how the distinct DDR in different tissues and cell types is regulated, due in part to the lack of animal models. We suggest that Zfas1 mice is a good model for studying the mechanism underlying tissue-specificity of DDR.

Actually, as a multi-systemic disorder, renal disease is highly prevalent in Cockayne syndrome patients(Stern-Delfils et al., 2020). Of interest, a recent study demonstrates that injured kidney cells can induce profound effects beyond their regular microenvironments, specifically on the central nervous system, showcasing the kidney as an active participant in complex inter-organ signaling(Mulderrig et al., 2021). In this study, we find that mice deficient in ZFAS1 impedes DNA damage repair, with marked physiological consequences in kidney, suggesting that kidney proximal tubule cells might be common target of DNA damage-induced transcriptional stress. However, the precise mechanisms by which ZFAS1 deficiency provokes kidney abnormalities through direct or indirect effects need to be further explored.

Taken together, mechanistically, we propose that the functions of ZFAS1 in DNA damage response are twofold: (1) to facilitate the exchange and retention of different isoforms of RNAPIIo to regulate transcription shutdown and recovery; (2) be complementary with but not limited to NME1 to regulate the initial events of DDR, such as dephosphorylation of γH2AX, to promote DNA repair and cell cycle recovery.

Human cell lines and culture conditions

SV40-immortalized human fetal lung fibroblast cells MRC5_VA were maintained in high glucose DMEM medium (Thermo Fisher Scientific, 11965118) supplemented with 10% v/v fetal bovine serum (PAN-Biotech, P30-3302) and 5% penicillin/streptomycin (Thermo Fisher Scientific, 15140122) at 37°C with 5% CO₂ and routinely passaged 2–3 times a week. Cell lines were confirmed to be mycoplasma-free.

Animals

Animal experiments undertaken in this study were with approval of the Guangzhou Medical University Animal Care and Use Committee. All mice were maintained under specific pathogen-free conditions in individually ventilated cages at 19–23°C with light from 07:00 to 19:00.

CRISPR/Cas9-mediated genome engineering was conducted to generate a Zfas1^−/− knockout C57BL/6 mouse. Cas9 vector and guide-RNA (target sequence: gRNA-A, 5’-GGGGTGGACCTCTGTACTATAGG-3’; gRNA-B, 5’-ACCATGCTCAAGTTGCTGCAAGG-3’) for mouse Zfas1 3’UTR were generated by in vitro transcription, followed by injected into fertilized eggs for knockout mouse production. The positive founders were genotyped by PCR and DNA sequencing analysis. The sequences of the primers used for genotyping were depicted in Table S11.

Acute effects in the skin of shaven Zfas1^−/− mice following exposure to UV-B at dose of 500 J/m²/day during 4 consecutive days were conducted by killing three mice per genotype 3 days after the last treatment. Skin samples were routinely processed (hematoxylin-eosin staining) for histopathologic examination as described before(Xu et al., 2019).

Western blotting

For whole cell extracts, cells pellets were lysed in NP-40 lysis buffer [50 mM Tris-HCl pH 7.5, 500 mM NaCl, 2 mM EDTA, 0.5% (v/v) NP-40, 0.5 mM DTT, and Protease Inhibitor Cocktail (made in house)]. Protein samples were resolved by 10% or 15% SDS-PAGE gels and transferred to PVDF membrane (GE healthcare Life Sciences, 10600023), followed by blocking in 5% (w/v) skimmed milk in TBST (50 mM Tris HCl pH 7.6, 150 mM NaCl, 0.1% (v/v) Tween 20) for 1 h at room temperature and incubated with primary antibody (in 5% (w/v) skimmed milk in TBST) overnight at 4°C. Antibody against tubulin or GAPDH served as loading controls. Membranes were washed three times in TBST, followed by incubation with 1:1000 diluted HRP-conjugated secondary antibodies in 5% (w/v) skimmed milk in TBST. After extensive washing with TBST, the proteins were visualized using Chemiluminescent Substrate ECL reagent (Thermo Fisher Scientific, 34075).

CPD and γH2A.X immunofluorescence

Cells were maintained on coverslip and treated with UV-C (10 J/m²) and recovered for the indicated period. After fixing in 4% paraformaldehyde for 15 min, cells were permeabilized in PBS-T (PBS plus 0.3% (v/v) Triton X-100), and the DNA was denatured in 2 M HCl for 5 min (for CPD immunofluorescence). Nonspecific binding was blocked in PBS-T (PBS plus 0.3% (v/v) Triton X-100) with 10% skimmed milk. The anti-CPD antibody (Cosmo Bio, CAC-NM-DND-001) or anti-γH2A.X antibody (Abcam, ab22551) was then applied in blocking solution (1:1000) and incubated at 4°C overnight. Secondary anti-mouse antibody conjugated to Alexa-fluor-488 was added to the coverslip for 1 h at room temperature. The coverslips were mounted onto slides using DAPI (Vector Laboratories, Inc. Peterborough, UK) and imaged as previously described(Liu et al., 2022).

Reverse transcriptase quantitative PCR

Total RNA was extracted using the RNeasy kit (QIAGEN, 74104) for nascent and mature RNA, following the instructions of the manufacturer including an on-column DNase digestion (QIAGEN, 79254). Reverse transcription was performed using PrimeScript™ RT Reagent Kit with gDNA Eraser (Takara, RR047A). cDNA was amplified using iQ SYBR green Supermix (Bio-Rad, 1708880) with 30 cycles of 15 s denaturation at 94°C, 15 s annealing at 60°C, and 20 s extensions at 72°C. Primer sequences are listed in Table S11. Unless otherwise noted, reference gene (GAPDH) normalized RNA expression was compared between variable (ex. UV-treated) and control samples using the Livak equation described before(Livak and Schmittgen, 2001).

Cell synchronization and cell cycle profile analysis

For all G1/S synchronization with thymidine, a double thymidine block was used as follows: cells were incubated for 14 h with 2 mM thymidine (Sigma-Aldrich, T1895), released for 9 h in fresh growth medium supplemented with 24 µM deoxycytidine (Sigma-Aldrich, D3897) after washing out the thymidine, and then blocked again with 2 mM thymidine for 14 h to arrest all cells at the beginning of S-phase. Cell cycle arrest was subsequently released with two washes of thymidine-free medium. After release, cells were treated with UV-C irradiation (10 J/m²) and harvested at the indicated time intervals.

Cell cycle analysis was carried out by flow cytometry. Briefly, cells were seeded onto 10-cm culture dishes, treated as indicated, collected, fixed, and stained with 25 µg/mL propidium iodide (Sigma-Aldrich, 81845) and 10 µg/mL RNAse A (Sigma-Aldrich, RNASEA-RO) in PBS for 30 mins at 37°C, acquired on a FACSCalibur flow cytometer instrument and analyzed using FACStation and FlowJo software (BD Biosciences). A minimum of ten thousand events were analyzed for each sample.

RNA-seq library construction and sequencing

G1/S-phase-synchronized MRC5_VA cells were either left untreated or treated with 10 J/m² of UV-C irradiation prior to release, followed by recovery for the indicated time periods from 3 to 96 hr. Total RNA was extracted and purified using RNeasy Mini Kit (QIAGEN, 74104) and analyzed on a 2100 Bioanalyzer (Agilent Technologies). All samples had an RIN (RNA Integrity Number) value of greater than 9.6. Subsequently, ribosome RNA was removed from the purified total RNA using Epicentre Ribo-Zero™ rRNA Removal Kit (Epicentre, MRZH11124), and the remaining RNA was used for library preparation using TruSeq Stranded Total RNA Sample Prep Kit (Illumina, 20020597). Libraries were quantified fluorometrically using Qubit® dsDNA HS Assay Kits (Thermo Fisher Scientific, Q32851) on a Qubit 2.0 Fluorometer (Thermo Fisher Scientific). All libraries (> 2 nM/µl) were sequenced on an Illumina NovaSeq 6000 (PE150 run) according to the manufacturer’s instruction to a depth of 220–340 million reads on average.

The sequence data were deposited in the NCBI Gene Expression Omnibus (GEO), with BioProject accession number, GSE239617.

Processing, analysis and graphic display of RNA-seq data

Raw reads were pre-processed with sequence grooming tool FASTQC followed by sequence alignment using HISAT2(Kim et al., 2015). Transcript levels were quantified as fragments per kilobase of transcript per million mapped reads (FPKM) generated by TopHat/Cufflinks(Kim et al., 2013). SAMtools(Li et al., 2009) were used to convert sam files to bam files in order to make raw data visible on IGV (Integrative Genomics Viewer) or UCSC Genome Browser. Transcript levels were converted to the log-space by taking the logarithm to the base 2. R studio (ggplot2 and gplots packages) was used to run custom R scripts to perform box plots, principal component analysis (PCA), dendrograms, and heatmaps. Samples were analyzed through DESeq2(Varet et al., 2016) to obtain log2 fold change and its respective p value. Differentially expressed transcripts have been identified on these transformed values by using the criteria of log2 (Fold Change) ≥ 1 and padj < 0.05. Gene ontology analysis of DEGs (differentially expressed genes) was performed using clusterProfiler(Yu et al., 2012).

Identification of novel lncRNAs from RNA-seq

The known noncoding RNAs expressed in at least one sample were identified by blasting the transcripts against the NONCODE v6.0 database(Zhao et al., 2021) using the following selection criteria: identity > 0.9, coverage > 0.8, and E-value < 10⁵. These transcripts were named as the ID number in the NONCODE v6.0 database.

Based on the features of lncRNA, a series of stringent screening conditions were established to identify novel lncRNAs from the RNA-seq data: (1) Cuffmerge was used to merge the transcripts which were obtained by splicing and remove transcripts with uncertain directions and < 200 bp in length; (2) Cuffcompare was used to filter out transcripts that overlap with the database annotation exon region; (3) Cuffquant was applied to calculate the expression level of each transcript and transcripts with FPKM ≥ 0.2 were selected; (4) Four analysis tools, including CNCI (Coding-Non-Coding-Index, v2), CPC (encoding potential calculator, 0.9-r2), Pfam Scan (v1.3) and phyloCSF were used to predict the coding potential of the transcripts.

ZFAS1 cloning and sequencing analysis (RACE)

3’ and 5’ RACE was performed using the SMARTer® RACE 5’/3’ Kit (Takara, 634858). RNA was extracted from either untreated or 10 J/m² UV-C exposed MRC5_VA cells, and RACE was performed according to the standard manufacturer’s protocol. Primers used for RACE assay were depicted in Table S11.

Plasmid construction

For construction of the ZFAS1-FLAG plasmids, the predicted ORF of ZFAS1, full length of ZFAS1, and the CDS of ACTB gene were cloned into the Sgfl and Mlul sites of the eukaryotic expression vector pCMV6-Entry with N-terminal start codon ATG and C-terminal Flag tag. Primers used for the cloning were depicted in Table S11.

RNA fluorescence in situ hybridization (FISH)

To detect ZFAS1 RNA, FISH analysis of MRC5_VA cells maintained on sterile glass coverslips was performed with a fluorescence in situ hybridization kit (RiboBio, C10910). Briefly, cells were rinsed in PBS and then fixed in 3.7% formaldehyde plus 10% acetic acid in PBS (pH 7.4) for 15 min at room temperature. Cells were subsequently permeabilized in PBS containing 0.2%-0.5% Triton X-100 and 5 mM vanadyl ribonucleoside complex (10 mM) (NEB, S1402) on ice for 5 min, washed in PBS 3 x 10 min and rinsed once in 2 x SSC buffer (Thermo Fisher Scientific, AM9770). Hybridization was carried out using oligodeoxynucleotide probes for ZFAS1, U6, and 18S (RiboBio) in a moist chamber at 37°C for 12–16 hours while protected from light according to the protocol for adherent mammalian cell lines. The next day, cells were counterstained with DAPI. Images were acquired on a laser scanning confocal microscope (Leica, TCS SP8).

Chromatin/RNA immunoprecipitation combined with quantitative PCR (ChIP-qPCR/RIP-qPCR)

MRC5_VA cells were plated onto 15-cm dishes at a density of 5 x 10⁶ cells per dish and treated with 10 J/m² UV-C exposure. After 3 hr of treatment, chromatin immunoprecipitation (ChIP) assays were carried out as described before(Wang et al., 2014). For each ChIP assay, the following antibodies (2 µg) were used: RNA polymerase II (Abcam, ab26721), H3K4me3 (Abcam, ab8580), H3K27ac (Abcam, ab4729), or immunoglobulin G control (Abcam, ab172730). ChIP enriched DNA and input DNA were subjected to qRT-PCR analysis with SYBR Green Supermix (Bio-Rad, 1708880). Enrichment by ChIP assay on the specific genomic regions was assessed relative to the immunoglobulin G control.

For RIP assay, SUPERase-In™ RNase inhibitor (1000 U/ml, Ambion, AM2694) and protease inhibitor were added into cell lysis buffer, and Ribonucleoside Vanadyl Complex (10 mM, NEB, S1402) was added into washing buffer. Antibodies specific to human NME1 (Proteintech, 11086-2-AP), SSRP1 (BioLegend, 609701) and XRCC5 (Proteintech, 16389-1-AP) were applied in the assay.

The primer sets were listed in Table S11.

6mA-IP-qPCR

Genomic DNA was isolated from untreated or UV-irradiated MRC5_VA cells and treated with RNase to remove RNA contamination. The DNA sample was then sonicated to produce 200–400 bp fragments. The fragmented DNA was incubated with specific anti-6mA antibody (Synaptic Systems, 202003) in immunoprecipitation buffer (2.7 mM KCl, 2 mM potassium phosphate, 137 mM NaCl, 10 mM Sodium phosphate, 0.05% Triton X-100, pH 7.4) for 2 h at 4°C. The mixture was then immunoprecipitated by Protein A/G Plus-Agarose (Santa Cruz, sc-2003) that was pre-blocked with bovine serum albumin (BSA) at 4°C for 2 h. After extensive washing, the bound DNA was eluted from the beads in elution buffer (50 mM Tris-HCl, pH 8.0, 1% SDS, 10 mM EDTA) at 65°C, treated with proteinase K and purified using QIAquick PCR Purification Kit (Qiagen, 28104). Fold-enrichment of each fragment was determined by quantitative real-time PCR. The primers used for 6mA-IP-qPCR assays were depicted in Table S11.

Clonogenic assay

Cultured MRC5_VA cells depleted with ZFAS1 or overexpressing ZFAS1 were treated with Trypsin/EDTA, counted and seeded onto 6-well plates at a low density (200–500 cells/well) for colony formation. Cells were incubated overnight followed by UV-C irradiation under the dose of 2, 5 and 10 J/m². Colonies were allowed to form over 14–21-day period and were stained with crystal violet and counted.

ZFAS1 knockdown and overexpression

The target sequences of ZFAS1, shared by both short and long transcript variants, were 5’- GCAGACATCTACAACCTTC − 3’ for shZFAS1-A, 5’- ATGGATTTTGGAAGAGGGA − 3’ for shZFAS1-B, and 5’- GAGTTTAAAAGGCTGTGCC − 3’ for shZFAS1-C (shZFAS1-C was used in the following functional assays). These sequences were cloned into lentiviral shRNA expression hCMV-SMARTvector.

For overexpression assays in MRC5_VA cells, the Lipofectamine 3000 (Invitrogen) was used according to the manufacturer’s protocol. The following plasmids were used: pCMV6-Entry-human ZFAS1 and pCMV6-Entry.

Alkaline comet assay

Comet assay was performed according to ENZO comet assay kit (Enzo life sciences, ADI-900-166) protocol for alkaline comet assay of adherent cells. Samples were visualized using a confocal microscope (Leica, TCS SP8) with 40 x magnification. Comet parameters were analyzed automatically with image acquisition and analysis software package CaspLab.

EdU staining

Wild-type or ZFAS1-depleted MRC5_VA cells were cultured on coverslips and washed with PBS once, followed by irradiation with 10 J/m² UV-C. After UV-C treatment, cells were immediately incubated with serum-free fresh medium supplemented with 10 µM EdU for 2 h. Cells were then washed with PBS once, followed by fixation with 3.7% formaldehyde for 15 minutes and permeabilization with PBS containing 0.5% Triton X-100 for 20 minutes. After extensive washing with PBS, cells were blocked with 10% FBS in PBS for 30 min. Incorporated EdU was detected by fluorescent-azide coupling reaction (Click-iT™ EdU cell proliferation kit, Invitrogen, C10337). Photograms of the cells were captured with a confocal microscope (Leica, TCS SP8). Captured images were processed and analyzed with ImageJ Software. At least 200 non-S-phase cells were randomly selected per sample. Data points presented in the text are the averages of intensities.

5’ Ethynyl Uridine Staining

EU staining to detect newly synthesized RNA was performed according to the manufacturer’s instructions (Click-iT™ RNA imaging Kits, Invitrogen, C10329). The wild-type or ZFAS1-depleted MRC5_VA cells were exposed to 10 J/m² UV-C irradiation and incubated for the indicated period of time. Media was replaced with fresh media containing 0.75 mM 5’ Ethynyl Uridine (EU) and cells were incubated for another 2 hrs. EU-containing media was then removed and cells were fixed in PBS buffered formaldehyde (3.7%) for 45 mins at room temperature, washed once with PBS and followed by permeabilization with 0.5% Triton X-100 diluted in PBS for 30 mins. Cells were washed once with PBS and then Alexa Fluor 488 Azide fluorophores were covalently attached to EU-containing RNA by click reaction for 1 hr at room temperature. Cells were then counterstained and mounted with mounting medium containing DAPI. Automated image acquisition of at least 5 fields per well was performed (Leica, TCS SP8).

Chromatin immunoprecipitation and NGS sequencing (ChIP-Seq)

RNAPII-CTD ChIP-Seq was performed under following conditions: the wild-type and ZFAS1-depleted MRC5_VA cells were UV irradiated (10 J/m² UV-C), followed by incubation for designated time intervals (3 hr, 6 hr, and 12 hr). Cells were harvested by trypsin-treatment and fixed in suspension with formaldehyde (1% final concentration) for 15 min at room temperature, with rotation. The crosslinking reaction was quenched with glycine (125 mM final concentration) for 5 min. Cells were then washed twice with ice-cold 1 x PBS, suspended in 1 ml of ChIP cell lysis buffer (5 mM HEPES pH 8.0, 85 mM KCl, 0.5% NP-40, and protease inhibitors) and incubated on ice for 5 min. Nuclei were pelleted by centrifugation at 3,900 x g for 5 min at 4°C, followed by suspension in ChIP nuclear lysis buffer (50 mM Tris-HCl pH 8.1, 10 mM EDTA pH 8.0, 1% SDS, and protease inhibitors) and incubated for 5 min on ice. Nuclear lysate was sheared by using an ice-water bath-embedded Bioruptor sonication system at high power, 30 sec on, 30 sec off mode for 5–10 min. The size of the sheared DNA was checked by 2% agarose gel electrophoresis to be between 300–500 base pairs (bp). Sonicated chromatin was cleared by centrifugation at 20,000 x g for 15 min at 4°C. Before the immune-precipitation, chromatin was diluted 1:5 with ChIP dilution buffer (0.01% SDS, 1.15 Triton X-100, 1.2 mM EDTA pH 8.0, 16.7 mM Tris-HCl pH 8.1, 167 mM NaCl, and protease inhibitors). 1 µg of anti-RPB1 phospho-Ser5-CTD (RPB1-Ser5, clone 3E8, Millipore, 04-1572-I), phospho-Ser2-CTD (RPB1-Ser2, clone 3E10, Millipore, 04-1571), or rat IgG (Invitrogen), was bound to 15 µl of Protein A/G agarose beads (Invitrogen) in 200 µl 5% BSA in 1 x PBS for 1 hr, before being washed twice with 500 µl of the same buffer. The sonicated chromatin was incubated with the antibody-conjugated beads overnight at 4°C with gentle rotation. Beads were washed twice with 1 ml of each of the following buffers: ChIP low salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.1, 150 mM NaCl); ChIP high salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.1, 500 mM NaCl); and ChIP LiCl buffer (10 mM Tris-HCl pH 8.0, 250 mM LiCl, 1% NP-40, 1% deoxycholic acid, and 1 mM EDTA). Beads were washed once with 1 ml of TE buffer (pH 8.0) and centrifuged for 5 min at 14,000 x g before removing the buffer. Beads were finally suspended in 40 µl Elution buffer (50 mM Tris-HCl pH 8.0, 10 mM EDTA, 1% SDS) and incubated at 65°C for 15 min. The eluted ChIP material was incubated at 65°C overnight to reverse the crosslinking with an additional 90 µl of 1% SDS in 1 x TE buffer and 1 µl of 10 mg/ml RNase A. In parallel, the input sample was also RNase treated and reverse crosslinked overnight at 65°C. Proteinase K (100 µg) and Glycogen (20 µg) were added to the eluted ChIP material and incubated for 2 hr at 37°C, and DNA was extracted by column purification (MinElute, QIAGEN, 28006). 1 ng of ChIPed DNA samples were submitted for further manipulation by standard ChIP-Seq library preparation techniques (NEBNext® Ultra™ DNA Library prep kit, NEB, #E7103) and sequenced under Illumina NovaSeq 6000 system, resulting in the production of 150 bp paired-end reads.

Alignment of the ChIP-seq reads was performed as follows: low-quality sequence reads and adapters were filtered out by Trimmomatic (v3.36)(Bolger et al., 2014). The trimmed reads were then aligned to the human reference genome (hg38) with the Burrows-Wheeler Aligner (BWA-v0.7.12-r1039)(Li, 2014). Duplicate reads were removed from the aligned reads by Biobambam2 (v2.0.72)(Tischler and Leonard, 2014). The mapped reads were visualized with the Integrated Genome Viewer (IGV-v2.3.90)(Thorvaldsdóttir et al., 2013). Further analysis was conducted using Bioconductor(Gentleman et al., 2004).

The sequence data were deposited in the NCBI Gene Expression Omnibus (GEO), with BioProject accession number, GSE239617.

ChIP-Seq TSS profiles

From the aligned ChIP-Seq reads, heatmaps and average density profiles were generated to draw spatial distributions of RPB1 along gene bodies under different conditions. Read densities at genomic regions (TSSs) of interest were extracted by SeqMINER 1.3.3(Ye et al., 2011). Heatmaps were generated directly in the software from matrixes of binned read densities for all considered individual items.

For determination of fold changes profiles, the number of reads per bin for a given sample were divided by the number of reads in the indicated control sample and expressed as Log₂ FC.

Promoter Escape Indexes

As described before(Chen et al., 2015), promoter escape indexes were calculated by taking the average coverage in rpm in the gene body (ranged from 101 bp to 2 kb downstream of TSS or 101 bp downstream of TSS to TTS for genes larger or smaller than 2 kb, respectively) divided by the average coverage on the promoter-proximal region.

RPB1 Peak Calling

Peaks were called against a rat IgG control using MACS v1.4.2(Zhang et al., 2008). All the peaks were listed in Table S7.

ChIRP-seq/ChIRP-MS

ChIRP was performed using biotinylated probes against ZFAS1 according to previously described methods(Chu et al., 2011). See Table S11 for probe sequences. Independent even and odd probe pools were used to ensure ZFAS1-specific retrieval. High-throughput sequencing libraries were constructed from ChIRPed DNA according to ChIP-seq protocol as described above, and sequenced on Illumina NovaSeq 6000 system with read length of 150bp. Raw reads were uniquely mapped to reference genome (hg38) using Bowtie(Langmead and Salzberg, 2012).

ZFAS1 peak calling, motif finding, and GO term analysis were conducted as described previously(Chu et al., 2011). The sequence data were deposited in the NCBI Gene Expression Omnibus (GEO), with BioProject accession number, GSE239617.

For ChIRP-MS, protein elution, extraction protein sample preparation for MS were conducted as described before(Chu et al., 2015). See Table S11 for ZFAS1 ChIRP probe design.

Behavioral studies

All behavioral studies were conducted during the light period. Mice were habituated to the procedure room for 1 hr before each test. In order to recover, mice were given three days between tests. Between subjects, all behavior apparatuses were sanitized with 70% ethanol solution and dried prior to introducing the next subject. All behavioral tests were performed with 8-week-old mice. Data are presented as means ± SD.

Open field test

Exploratory behavior in a novel environment was assessed by a 30 min session in an open field chamber (45cmL x 45cmW x 30cmH). Mice were placed in the middle of the central area and then recording was started. Locomotor activity (distance travelled and velocity) in the center/periphery of the arena were recorded by using a video camera and analyzed using the EthoVisione XT software (Noldus).

Morris water maze test

Mice were placed in a circular pool (120 cm in diameter and 50 cm height) containing opacified water. On the first day of training, before the first trial, mice were placed on the platform for 30 sec, followed by a 30-sec practice of swimming and three practice climbs onto the platform. The mouse was then placed into the water facing the wall of the pool and allowed to search for the platform. The trial ended when the mice climbed onto the platform or when a maximum of 60 sec elapsed. At the end of each trial, the mouse was allowed to rest on the platform for 60 sec. Then mice were immediately placed into the water again and released to swim and find the platform from a different start location. Four consecutive trials were administered for each animal. On day 5, the evacuation platform was removed and probing was performed.

Isolation of primary mouse lung fibroblasts

Newborn mice were anesthetized with aether, the whole lung was removed. Approximately 1 cm² fragment of the tissue was excised and placed immediately in sterile PBS to avoid drying. The tissue was cut into 1 mm pieces, transferred into a sterile 30 ml beaker containing 10 ml of DMEM/F12, 0.14 Wunsch units/ml liberase blendzyme 3 and 1 x antibiotic/antimycotic, incubated at 37°C for 30–90 minutes with stirring. After digestion, the tissue fragments were rinsed 3 times with 10 ml of warm DMEM/F12 media containing 15% FBS and 1 x antibiotic/antimycotic, and then centrifuged at 524 x g. The pellet was resuspended in 10 ml of warm DMEM/F12 media with 15% FBS, 1 x antibiotic/antimycotic. The tissue pieces were broken down by pipetting with maximum force. Another 30 ml of DMEM/F12 media with 15% FBS, 1 x antibiotic/antimycotic was added and the solution was centrifuged twice at 524 x g to ensure removal of Liberase. The cell pellets were resuspended in 10 ml of DMEM/F12 media with 15% FBS, 1 x antibiotic/antimycotic and transferred to a 10 cm tissue culture dish and placed in a tissue culture incubator at 37°C, 5% CO₂.

Immunohistochemistry

Mice were anesthetized with aether, and subjected to cardiac perfusion with saline, followed by a 10% formalin flush. Tissues (kidney or brain) were removed and sectioned into 3 mm slices before transfer into formalin. Tissues were fixed in 10% formalin for a minimum of 48 h at room temperature and then subjected to paraffin embedding schedule as follow: 70% Ethanol, two changes,1 h each; 80% Ethanol, one change, 1 h; 95% Ethanol, one change, 1 h; 100% Ethanol, three changes, 1.5 h each; Xylene, three changes, 1.5 h each; Paraffin wax (58°C-60°C), two changes, 2 h each. After the paraffin wax cooled down and solidified, the paraffin blocks were trimmed and cut at 5 mm and placed in water bath at about 40°C-45°C. Sections were mounted into slides and air-dried for 30 minutes, then baked in 45°C oven overnight. Before deparaffinization, slides were baked in 65°C oven for 2 h. Slides were placed in a rack and subjected to deparaffinization and dehydration in the following washes: Xylene, two changes, 3 min each; Xylene 1:1 with 100% ethanol, one change, 3 min; 100% ethanol, two changes, 3 min each; 95% ethanol, one change, 3 min; 70% ethanol, one change, 3 min; 50% ethanol, one change, 3 min. Slides were kept in the tap water until ready to perform antigen retrieval. Slides were placed to a boil in antigen retrieval buffer (10 mM Sodium Citrate, 0.05% Tween20, pH 6.0), then maintained at a sub-boiling temperature for 10 min. Slides were cooled down in running tap water for 5 min. For H&E staining, slides were stained in hematoxylin for 3–5 min before antigen retrieval, and then washed in running tap water until sections blue for 5 min or less. Slides were differentiated in 1% acid alcohol (1% HCl in 70% alcohol) for 5 min and then washed in running tap water until the sections were again blue by dipping in an alkaline solution (ammonia water) followed by tap water wash. Slides were stained in 1% Eosin for 10 min and washed in tap water for 1–5 min. Slides were dehydrated through 95% alcohol, 2 changes of absolute alcohol, 5 min each, and cleared in 2 changes of Xylene, 5 min each. Finally, slides were mounted in mounting media. For cell-specific immunohistochemical staining, slides were blocked in blocking solution (1 X PBS-T containing 10% FBS, 1% BSA and 0.3% Triton X-100) for 1 h at room temperature after antigen retrieval. Slides were incubated in primary antibodies (diluted in 1 X PBS-T containing 1% BSA) overnight at 4°C. Slides were washed three times in PBS-T and incubated in fluorophore-conjugated secondary antibody diluted in PBS-T, 1% BSA for 1 h at room temperature. Slides were washed three times in PBS-T and mounted with mounting medium with DAPI.

Quantification and statistical analysis

A two-tailed Student’s t-Test was used for analysis of statistical significance, with a P < 0.05 considered significant. For mouse studies, we determined that a sample size of at least 3–6 animals per genotype per sex was required.

Data availability

Genome-wide data used in this study are available under GEO number GSE239617.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (grant number 82271195 and 32100435) to Y.W. We thank Novogene sequencing facility for their expert assistance.

Author contributions

Y.W. conceived the project. J.L., T.Z., Z.H., and Q.L. performed all experiments. N.H., Z.F. and Z.W. performed data analysis. Y.X. coordinated the study. Y.W. wrote the manuscript, with input from all authors.

Declaration of interests

The authors declare no competing interests.

Adam, S., Polo, S.E., and Almouzni, G. (2013). Transcription recovery after DNA damage requires chromatin priming by the H3.3 histone chaperone HIRA. Cell 155, 94–106.
Adar, S., Hu, J., Lieb, J.D., and Sancar, A. (2016). Genome-wide kinetics of DNA excision repair in relation to chromatin state and mutagenesis. Proc Natl Acad Sci U S A 113, E2124-2133.
Adriaens, C., Standaert, L., Barra, J., Latil, M., Verfaillie, A., Kalev, P., Boeckx, B., Wijnhoven, P.W., Radaelli, E., Vermi, W., et al. (2016). p53 induces formation of NEAT1 lncRNA-containing paraspeckles that modulate replication stress response and chemosensitivity. Nat Med 22, 861–868.
Andrade-Lima, L.C., Veloso, A., Paulsen, M.T., Menck, C.F., and Ljungman, M. (2015). DNA repair and recovery of RNA synthesis following exposure to ultraviolet light are delayed in long genes. Nucleic Acids Res 43, 2744–2756.
Anindya, R., Aygun, O., and Svejstrup, J.Q. (2007). Damage-induced ubiquitylation of human RNA polymerase II by the ubiquitin ligase Nedd4, but not Cockayne syndrome proteins or BRCA1. Molecular cell 28, 386–397.
Bartek, J., and Lukas, J. (2007). DNA damage checkpoints: from initiation to recovery or adaptation. Curr Opin Cell Biol 19, 238–245.
Bogliolo, M., and Surralles, J. (2015). Fanconi anemia: a model disease for studies on human genetics and advanced therapeutics. Curr Opin Genet Dev 33, 32–40.
Bolger, A.M., Lohse, M., and Usadel, B. (2014). Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30, 2114–2120.
Brueckner, F., Hennecke, U., Carell, T., and Cramer, P. (2007). CPD damage recognition by transcribing RNA polymerase II. Science (New York, NY) 315, 859–862.
Campos, A., and Clemente-Blanco, A. (2020). Cell Cycle and DNA Repair Regulation in the Damage Response: Protein Phosphatases Take Over the Reins. Int J Mol Sci 21.
Chao, H.X., Poovey, C.E., Privette, A.A., Grant, G.D., Chao, H.Y., Cook, J.G., and Purvis, J.E. (2017). Orchestration of DNA Damage Checkpoint Dynamics across the Human Cell Cycle. Cell Syst 5, 445–459 e445.
Chen, F.X., Woodfin, A.R., Gardini, A., Rickels, R.A., Marshall, S.A., Smith, E.R., Shiekhattar, R., and Shilatifard, A. (2015). PAF1, a Molecular Regulator of Promoter-Proximal Pausing by RNA Polymerase II. Cell 162, 1003–1015.
Choudhuri, T., Murakami, M., Kaul, R., Sahu, S.K., Mohanty, S., Verma, S.C., Kumar, P., and Robertson, E.S. (2010). Nm23-H1 can induce cell cycle arrest and apoptosis in B cells. Cancer Biol Ther 9, 1065–1078.
Chu, C., Qu, K., Zhong, F.L., Artandi, S.E., and Chang, H.Y. (2011). Genomic maps of long noncoding RNA occupancy reveal principles of RNA-chromatin interactions. Molecular cell 44, 667–678.
Chu, C., Zhang, Q.C., da Rocha, S.T., Flynn, R.A., Bharadwaj, M., Calabrese, J.M., Magnuson, T., Heard, E., and Chang, H.Y. (2015). Systematic discovery of Xist RNA binding proteins. Cell 161, 404–416.
Clark, M.B., Johnston, R.L., Inostroza-Ponta, M., Fox, A.H., Fortini, E., Moscato, P., Dinger, M.E., and Mattick, J.S. (2012). Genome-wide analysis of long noncoding RNA stability. Genome Res 22, 885–898.
Cook, P.J., Ju, B.G., Telese, F., Wang, X., Glass, C.K., and Rosenfeld, M.G. (2009). Tyrosine dephosphorylation of H2AX modulates apoptosis and survival decisions. Nature 458, 591–596.
Diaz-Lagares, A., Crujeiras, A.B., Lopez-Serra, P., Soler, M., Setien, F., Goyal, A., Sandoval, J., Hashimoto, Y., Martinez-Cardus, A., Gomez, A., et al. (2016). Epigenetic inactivation of the p53-induced long noncoding RNA TP53 target 1 in human cancer. Proc Natl Acad Sci U S A 113, E7535-E7544.
Dinant, C., Ampatziadis-Michailidis, G., Lans, H., Tresini, M., Lagarou, A., Grosbart, M., Theil, A.F., van Cappellen, W.A., Kimura, H., Bartek, J., et al. (2013). Enhanced chromatin dynamics by FACT promotes transcriptional restart after UV-induced DNA damage. Molecular cell 51, 469–479.
Epanchintsev, A., Costanzo, F., Rauschendorf, M.A., Caputo, M., Ye, T., Donnio, L.M., Proietti-de-Santis, L., Coin, F., Laugel, V., and Egly, J.M. (2017). Cockayne's Syndrome A and B Proteins Regulate Transcription Arrest after Genotoxic Stress by Promoting ATF3 Degradation. Molecular cell 68, 1054–1066 e1056.
Gaillard, H., and Aguilera, A. (2013). Transcription coupled repair at the interface between transcription elongation and mRNP biogenesis. Bba-Gene Regul Mech 1829, 141–150.
Gentleman, R.C., Carey, V.J., Bates, D.M., Bolstad, B., Dettling, M., Dudoit, S., Ellis, B., Gautier, L., Ge, Y., Gentry, J., et al. (2004). Bioconductor: open software development for computational biology and bioinformatics. Genome biology 5, R80.
Giglia-Mari, G., Zotter, A., and Vermeulen, W. (2011). DNA damage response. Cold Spring Harbor perspectives in biology 3, a000745.
Goldenson, B., and Crispino, J.D. (2015). The aurora kinases in cell cycle and leukemia. Oncogene 34, 537–545.
Gregersen, L.H., and Svejstrup, J.Q. (2018). The Cellular Response to Transcription-Blocking DNA Damage. Trends Biochem Sci 43, 327–341.
Han, A., Sinclair, W.K., and Yu, C.K. (1971). Ultraviolet light-induced division delay in synchronized Chinese hamster cells. Biophys J 11, 540–549.
Hanasoge, S., and Ljungman, M. (2007). H2AX phosphorylation after UV irradiation is triggered by DNA repair intermediates and is mediated by the ATR kinase. Carcinogenesis 28, 2298–2304.
Hanawalt, P.C., and Spivak, G. (2008). Transcription-coupled DNA repair: two decades of progress and surprises. Nat Rev Mol Cell Biol 9, 958–970.
Heine, G.F., Horwitz, A.A., and Parvin, J.D. (2008). Multiple mechanisms contribute to inhibit transcription in response to DNA damage. J Biol Chem 283, 9555–9561.
Helleday, T., Eshtad, S., and Nik-Zainal, S. (2014). Mechanisms underlying mutational signatures in human cancers. Nat Rev Genet 15, 585–598.
Hung, T., Wang, Y.L., Lin, M.F., Koegel, A.K., Kotake, Y., Grant, G.D., Horlings, H.M., Shah, N., Umbricht, C., Wang, P., et al. (2011). Extensive and coordinated transcription of noncoding RNAs within cell-cycle promoters. Nature genetics 43, 621-U196.
Jaarsma, D., van der Pluijm, I., van der Horst, G.T., and Hoeijmakers, J.H. (2013). Cockayne syndrome pathogenesis: lessons from mouse models. Mech Ageing Dev 134, 180–195.
Jackson, S.P., and Bartek, J. (2009). The DNA-damage response in human biology and disease. Nature 461, 1071–1078.
Jarrett, S.G., Novak, M., Dabernat, S., Daniel, J.Y., Mellon, I., Zhang, Q., Harris, N., Ciesielski, M.J., Fenstermaker, R.A., Kovacic, D., et al. (2012). Metastasis suppressor NM23-H1 promotes repair of UV-induced DNA damage and suppresses UV-induced melanomagenesis. Cancer Res 72, 133–143.
Ji, P., Agrawal, S., Diederichs, S., Bäumer, N., Becker, A., Cauvet, T., Kowski, S., Beger, C., Welte, K., Berdel, W.E., et al. (2005). Cyclin A1, the alternative A-type cyclin, contributes to G1/S cell cycle progression in somatic cells. Oncogene 24, 2739–2744.
Johnsson, P., Lipovich, L., Grandér, D., and Morris, K.V. (2014). Evolutionary conservation of long non-coding RNAs; sequence, structure, function. Biochim Biophys Acta 1840, 1063–1071.
Joukov, V., and De Nicolo, A. (2018). Aurora-PLK1 cascades as key signaling modules in the regulation of mitosis. Science signaling 11.
Kim, D., Langmead, B., and Salzberg, S.L. (2015). HISAT: a fast spliced aligner with low memory requirements. Nat Methods 12, 357–360.
Kim, D., Pertea, G., Trapnell, C., Pimentel, H., Kelley, R., and Salzberg, S.L. (2013). TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome biology 14, R36.
Kremer, B.E., Adang, L.A., and Macara, I.G. (2007). Septins regulate actin organization and cell-cycle arrest through nuclear accumulation of NCK mediated by SOCS7. Cell 130, 837–850.
Langmead, B., and Salzberg, S.L. (2012). Fast gapped-read alignment with Bowtie 2. Nat Methods 9, 357–359.
Lans, H., Hoeijmakers, J.H.J., Vermeulen, W., and Marteijn, J.A. (2019). The DNA damage response to transcription stress. Nat Rev Mol Cell Biol 20, 766–784.
Lavigne, M.D., Konstantopoulos, D., Ntakou-Zamplara, K.Z., Liakos, A., and Fousteri, M. (2017). Global unleashing of transcription elongation waves in response to genotoxic stress restricts somatic mutation rate. Nat Commun 8, 2076.
Le May, N., Mota-Fernandes, D., Velez-Cruz, R., Iltis, I., Biard, D., and Egly, J.M. (2010). NER factors are recruited to active promoters and facilitate chromatin modification for transcription in the absence of exogenous genotoxic attack. Molecular cell 38, 54–66.
Li, H. (2014). Toward better understanding of artifacts in variant calling from high-coverage samples. Bioinformatics 30, 2843–2851.
Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N., Marth, G., Abecasis, G., Durbin, R., and Genome Project Data Processing, S. (2009). The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079.
Liakos, A., Konstantopoulos, D., Lavigne, M.D., and Fousteri, M. (2020). Continuous transcription initiation guarantees robust repair of all transcribed genes and regulatory regions. Nat Commun 11, 916.
Limsirichaikul, S., Niimi, A., Fawcett, H., Lehmann, A., Yamashita, S., and Ogi, T. (2009). A rapid non-radioactive technique for measurement of repair synthesis in primary human fibroblasts by incorporation of ethynyl deoxyuridine (EdU). Nucleic Acids Res 37, e31.
Liu, J., Liu, L., He, J., Xu, Y., and Wang, Y. (2021). Multi-omic analysis of altered transcriptome and epigenetic signatures in the UV-induced DNA damage response. DNA repair 106, 103172.
Liu, J., Wu, Z., He, J., and Wang, Y. (2022). Cellular fractionation reveals transcriptome responses of human fibroblasts to UV-C irradiation. Cell death & disease 13, 177.
Livak, K.J., and Schmittgen, T.D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25, 402–408.
Ma, D., McCorkle, J.R., and Kaetzel, D.M. (2004). The metastasis suppressor NM23-H1 possesses 3'-5' exonuclease activity. J Biol Chem 279, 18073–18084.
Madabhushi, R., Pan, L., and Tsai, L.H. (2014). DNA damage and its links to neurodegeneration. Neuron 83, 266–282.
Mah, L.J., El-Osta, A., and Karagiannis, T.C. (2010). gammaH2AX: a sensitive molecular marker of DNA damage and repair. Leukemia 24, 679–686.
Mayne, L.V., and Lehmann, A.R. (1982). Failure of RNA synthesis to recover after UV irradiation: an early defect in cells from individuals with Cockayne's syndrome and xeroderma pigmentosum. Cancer Res 42, 1473–1478.
McKay, B.C., Stubbert, L.J., Fowler, C.C., Smith, J.M., Cardamore, R.A., and Spronck, J.C. (2004). Regulation of ultraviolet light-induced gene expression by gene size. Proc Natl Acad Sci U S A 101, 6582–6586.
Mellon, I., Spivak, G., and Hanawalt, P.C. (1987). Selective removal of transcription-blocking DNA damage from the transcribed strand of the mammalian DHFR gene. Cell 51, 241–249.
Mulderrig, L., Garaycoechea, J.I., Tuong, Z.K., Millington, C.L., Dingler, F.A., Ferdinand, J.R., Gaul, L., Tadross, J.A., Arends, M.J., O'Rahilly, S., et al. (2021). Aldehyde-driven transcriptional stress triggers an anorexic DNA damage response. Nature 600, 158–163.
Müllers, E., Silva Cascales, H., Jaiswal, H., Saurin, A.T., and Lindqvist, A. (2014). Nuclear translocation of Cyclin B1 marks the restriction point for terminal cell cycle exit in G2 phase. Cell cycle (Georgetown, Tex) 13, 2733–2743.
Nakazawa, Y., Hara, Y., Oka, Y., Komine, O., van den Heuvel, D., Guo, C., Daigaku, Y., Isono, M., He, Y., Shimada, M., et al. (2020). Ubiquitination of DNA Damage-Stalled RNAPII Promotes Transcription-Coupled Repair. Cell 180, 1228–1244 e1224.
Nyberg, K.A., Michelson, R.J., Putnam, C.W., and Weinert, T.A. (2002). Toward maintaining the genome: DNA damage and replication checkpoints. Annual review of genetics 36, 617–656.
O'Driscoll, M., Ruiz-Perez, V.L., Woods, C.G., Jeggo, P.A., and Goodship, J.A. (2003). A splicing mutation affecting expression of ataxia-telangiectasia and Rad3-related protein (ATR) results in Seckel syndrome. Nature genetics 33, 497–501.
Premi, S., Han, L., Mehta, S., Knight, J., Zhao, D., Palmatier, M.A., Kornacker, K., and Brash, D.E. (2019). Genomic sites hypersensitive to ultraviolet radiation. Proc Natl Acad Sci U S A 116, 24196–24205.
Proietti-De-Santis, L., Drane, P., and Egly, J.M. (2006). Cockayne syndrome B protein regulates the transcriptional program after UV irradiation. EMBO J 25, 1915–1923.
Rando, O.J., and Chang, H.Y. (2009). Genome-wide views of chromatin structure. Annu Rev Biochem 78, 245–271.
Rinn, J.L., and Chang, H.Y. (2012). Genome regulation by long noncoding RNAs. Annu Rev Biochem 81, 145–166.
Rockx, D.A., Mason, R., van Hoffen, A., Barton, M.C., Citterio, E., Bregman, D.B., van Zeeland, A.A., Vrieling, H., and Mullenders, L.H. (2000). UV-induced inhibition of transcription involves repression of transcription initiation and phosphorylation of RNA polymerase II. Proc Natl Acad Sci U S A 97, 10503–10508.
Sabarinathan, R., Mularoni, L., Deu-Pons, J., Gonzalez-Perez, A., and Lopez-Bigas, N. (2016). Nucleotide excision repair is impaired by binding of transcription factors to DNA. Nature 532, 264-+.
Sancar, A., Lindsey-Boltz, L.A., Unsal-Kacmaz, K., and Linn, S. (2004). Molecular mechanisms of mammalian DNA repair and the DNA damage checkpoints. Annu Rev Biochem 73, 39–85.
Saponaro, M., Kantidakis, T., Mitter, R., Kelly, G.P., Heron, M., Williams, H., Soding, J., Stewart, A., and Svejstrup, J.Q. (2014). RECQL5 controls transcript elongation and suppresses genome instability associated with transcription stress. Cell 157, 1037–1049.
Schmitt, A.M., Garcia, J.T., Hung, T., Flynn, R.A., Shen, Y., Qu, K., Payumo, A.Y., Peres-da-Silva, A., Broz, D.K., Baum, R., et al. (2016). An inducible long noncoding RNA amplifies DNA damage signaling. Nature genetics 48, 1370–1376.
Scruggs, B.S., Gilchrist, D.A., Nechaev, S., Muse, G.W., Burkholder, A., Fargo, D.C., and Adelman, K. (2015). Bidirectional Transcription Arises from Two Distinct Hubs of Transcription Factor Binding and Active Chromatin. Molecular cell 58, 1101–1112.
Shaltiel, I.A., Krenning, L., Bruinsma, W., and Medema, R.H. (2015). The same, only different - DNA damage checkpoints and their reversal throughout the cell cycle. J Cell Sci 128, 607–620.
Sogo, J.M., Lopes, M., and Foiani, M. (2002). Fork reversal and ssDNA accumulation at stalled replication forks owing to checkpoint defects. Science (New York, NY) 297, 599–602.
Stern-Delfils, A., Spitz, M.A., Durand, M., Obringer, C., Calmels, N., Olagne, J., Pillay, K., Fieggen, K., Laugel, V., and Zaloszyc, A. (2020). Renal disease in Cockayne syndrome. Eur J Med Genet 63, 103612.
Struhl, K. (2007). Transcriptional noise and the fidelity of initiation by RNA polymerase II. Nat Struct Mol Biol 14, 103–105.
Thorvaldsdóttir, H., Robinson, J.T., and Mesirov, J.P. (2013). Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Briefings in bioinformatics 14, 178–192.
Tischler, G., and Leonard, S. (2014). biobambam: tools for read pair collation based algorithms on BAM files. Source Code for Biology and Medicine 9, 13.
van der Horst, G.T., van Steeg, H., Berg, R.J., van Gool, A.J., de Wit, J., Weeda, G., Morreau, H., Beems, R.B., van Kreijl, C.F., de Gruijl, F.R., et al. (1997). Defective transcription-coupled repair in Cockayne syndrome B mice is associated with skin cancer predisposition. Cell 89, 425–435.
van Oosten, M., Rebel, H., Friedberg, E.C., van Steeg, H., van der Horst, G.T., van Kranen, H.J., Westerman, A., van Zeeland, A.A., Mullenders, L.H., and de Gruijl, F.R. (2000). Differential role of transcription-coupled repair in UVB-induced G2 arrest and apoptosis in mouse epidermis. Proc Natl Acad Sci U S A 97, 11268–11273.
Varet, H., Brillet-Gueguen, L., Coppee, J.Y., and Dillies, M.A. (2016). SARTools: A DESeq2- and EdgeR-Based R Pipeline for Comprehensive Differential Analysis of RNA-Seq Data. PLoS One 11, e0157022.
Vidakovic, A.T., Mitter, R., Kelly, G.P., Neumann, M., Harreman, M., Rodriguez-Martinez, M., Herlihy, A., Weems, J.C., Boeing, S., Encheva, V., et al. (2020). Regulation of the RNAPII Pool Is Integral to the DNA Damage Response. Cell 180, 1245-+.
Wang, Y., Chakravarty, P., Ranes, M., Kelly, G., Brooks, P.J., Neilan, E., Stewart, A., Schiavo, G., and Svejstrup, J.Q. (2014). Dysregulation of gene expression as a cause of Cockayne syndrome neurological disease. Proc Natl Acad Sci U S A 111, 14454–14459.
Wapinski, O., and Chang, H.Y. (2011). Long noncoding RNAs and human disease. Trends in cell biology 21, 354–361.
Williamson, L., Saponaro, M., Boeing, S., East, P., Mitter, R., Kantidakis, T., Kelly, G.P., Lobley, A., Walker, J., Spencer-Dene, B., et al. (2017). UV Irradiation Induces a Non-coding RNA that Functionally Opposes the Protein Encoded by the Same Gene. Cell 168, 843-+.
Wilson, M.D., Harreman, M., and Svejstrup, J.Q. (2013). Ubiquitylation and degradation of elongating RNA polymerase II: the last resort. Biochim Biophys Acta 1829, 151–157.
Wong, J.C., and Buchwald, M. (2002). Disease model: Fanconi anemia. Trends Mol Med 8, 139–142.
Xiao, C.L., Zhu, S., He, M., Chen, D., Zhang, Q., Chen, Y., Yu, G., Liu, J., Xie, S.Q., Luo, F., et al. (2018). N(6)-Methyladenine DNA Modification in the Human Genome. Molecular cell 71, 306–318.e307.
Xu, Y., Wu, Z., Liu, L., Liu, J., and Wang, Y. (2019). Rat Model of Cockayne Syndrome Neurological Disease. Cell Rep 29, 800–809.e805.
Ye, T., Krebs, A.R., Choukrallah, M.A., Keime, C., Plewniak, F., Davidson, I., and Tora, L. (2011). seqMINER: an integrated ChIP-seq data interpretation platform. Nucleic Acids Res 39, e35.
Yildirim, O., Izgu, E.C., Damle, M., Chalei, V., Ji, F., Sadreyev, R.I., Szostak, J.W., and Kingston, R.E. (2020). S-phase Enriched Non-coding RNAs Regulate Gene Expression and Cell Cycle Progression. Cell Rep 31, 107629.
Yu, G., Wang, L.G., Han, Y., and He, Q.Y. (2012). clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 16, 284–287.
Zhang, Y., Liu, T., Meyer, C.A., Eeckhoute, J., Johnson, D.S., Bernstein, B.E., Nusbaum, C., Myers, R.M., Brown, M., Li, W., et al. (2008). Model-based analysis of ChIP-Seq (MACS). Genome biology 9, R137.
Zhao, L., Wang, J., Li, Y., Song, T., Wu, Y., Fang, S., Bu, D., Li, H., Sun, L., Pei, D., et al. (2021). NONCODEV6: an updated database dedicated to long non-coding RNA annotation in both animals and plants. Nucleic Acids Res 49, D165-d171.

There is NO Competing Interest.

TableS1.xlsx
Dataset 1 Table S1. FPKM values for all the transcripts (including mRNAs and lncRNAs) detected in the study from RNA-seq Data in the synchronized MRC5_VA cells released from the G1/S phase for the indicated time intervals, Related to Figure 1.
TableS2.xlsx
Dataset 2 Table S2. Clustering and GO term Enrichment analysis of mRNAs for RNA-seq Data derived from non-irradiated G1/S-phase-synchronized MRC5_VA cells at the indicated intervals after release, Related to Figure S2A.
TableS3.xlsx
Dataset 3 Table S3. Clustering analysis of mRNAs for RNA-seq Data derived from UV-irradiated G1/S-phase-synchronized MRC5_VA cells at the indicated intervals after release, Related to Figure S2C.
TableS4.xlsx
Dataset 4 Table S4. GO term enrichment analysis of DEGs at 3 h and 72 h after UV-C irradiation, related to Figure S2C.
TableS5.xlsx
Dataset 5 Table S5. Clustering and co-localization Enrichment analysis of lncRNAs for RNA-seq data derived from the non-irradiated synchronized MRC5_VA cells released from the G1/S block, Related to Figure 2A.
TableS6.xlsx
Dataset 6 Table S6. Clustering and co-localization and co-expression Enrichment analysis of lncRNAs for RNA-seq data derived from the UV-irradiated synchronized MRC5_VA cells released from the G1/S block, Related to Figure 2B.
TableS7.xlsx
Dataset 7 Table S7. The list of peaks called by MACS from RNAPIIo ChIP-seq in the WT and ZFAS1-depleted cells treated with 10 J/m² UV-C irradiation, Related to Figure 5I.
TableS8.xlsx
Dataset 8 Table S8. The list of peaks (raw peaks) called by MACS from ZFAS1 ChIRP-seq in MRC5_VA cells treated with 10 J/m² UV irradiation, Related to Figure 6.
TableS9.xlsx
Dataset 9 Table S9. The list of peaks (true peaks) called by the stringent pipeline from ZFAS1 ChIRP-seq in MRC5_VA cells treated with 10 J/m² UV irradiation, Related to Figure 6.
TableS10.xlsx
Dataset 10 Table S10. The list of Gene Ontology (GO) terms of the genes with the identified filtered ZFAS1ChIRP-seq peaks (true peaks) in the wild type and UV-irradiated cells, Related to Figure 6.
TableS11.xlsx
Dataset 11 Table S11. List of oligos used in the study.
FigureS1.jpg
Extended Data 1 Figure S1. Characterization of cellular response upon UV-C irradiation, Related to Figure 1 (A) Immunofluorescence staining of CPD dimers in synchronized MRC5_VA cells released for the indicated time intervals from the G1/S block after UV-C treatment (10 J/m²); nuclei were visualized by DAPI staining. Scale bar = 100 μm. (B) Nascent mRNA production in two distinct regions of the human KIFAP3 gene following UV damage in synchronized MRC5_VA cells released from the G1/S block. Means ±SD are shown from three independent experiments. (C) Pie charts showing the number of different types of transcripts detected in the study. (D) PCA analysis of temporal transcriptomic data for mRNAs and lncRNAs respectively.
FigureS2.jpg
Extended Data 2 Figure S2. Stage-specific protein-coding gene expression profiles of human fibroblasts with or without genotoxic stress, Related to Figure 1 (A) Heatmap of transformed scaled read counts of temporal genes showing data from RNA-seq of non-irradiated MRC5_VA cells released from the G1/S block. Transcript-wise hierarchical clustering heatmap of differentially expressed genes showing segregation into six groups. Cluster 2 (n = 3,764) includes genes with increased expression at 3 h after release. Cluster 3 (n = 2,206) includes genes upregulated at 6 h after release. Plots show a representative expression profile for each indicated cluster. Enriched GO terms are shown to the right. (B) Box plots illustrating expression of representative marker genes in the indicated clusters shown in (A). (C) Heatmap of transformed scaled read counts of temporal genes showing data from RNA-seq of UV-irradiated MRC5_VA cells released from the G1/S block. Transcript-wise hierarchical clustering heatmap of differentially expressed genes showing segregation into seven groups. Cluster 2 (n = 725) includes genes with increased expression at 3 h after release. Cluster 6 (n = 1,382) includes genes upregulated at 72 h after release. Plots show a representative expression profile for each highlighted cluster. Enriched GO terms are shown to the right.
FigureS3.jpg
Extended Data 3 Figure S3. Expression changes of cell cycle regulators across DDR, Related to Figure 1 (A) Expression changes of two individual cyclin markers in MRC5_VA cells across the cell cycle with or without treatment of UV-C irradiation (10 J/m²). (B) Expression changes of two individual spindle genes in MRC5_VA cells across the cell cycle with or without treatment of UV-C irradiation (10 J/m²). (C) Expression changes of two individual transcription repressors in MRC5_VA cells across the cell cycle with or without treatment of UV-C irradiation (10 J/m²).
FigureS4.jpg
Extended Data 4 Figure S4. Features of DNA damage induced lncRNAs, Related to Figure 2 (A)-(B) Changes of two groups of representative lncRNAs across the cell cycle in the absence or presence of genotoxic stress. (C) Exon number for all lncRNAs in the genome and the DNA damage-induced lncRNAs that are grouped into the indicated clusters. (D) Distribution of CDS sizes for the DNA damage-induced lncRNAs grouped into the indicated clusters. Each bar represents the number of CDSs in each size class in increments of 80 bp. Where there are several CDS isoforms in one gene, they are all counted. (E) Transcript length between all lncRNAs and the DNA damage-induced lncRNAs grouped into the indicated clusters.
FigureS5.jpg
Extended Data 5 Figure S5. Identification of ZFAS1 as a DNA damage-induced lncRNA, Related to Figure 3 (A) Real-time PCR validations of SNHG15, SLC25A34-AS1, and ZFAS1 in the G1/S-phase-synchronized MRC5_VA cells upon UV-C irradiation. The expression of each gene was normalized to house-keeping gene GAPDH and presented relative to non-irradiated samples. (B) RNA-seq expression of three distinct ZFAS1 isoforms in the G1/S-phase-synchronized MRC5_VA cells during the DNA damage-induced cell cycle progression. (C) IGV snapshot showing transcription across the protein-coding gene ZNFX1 locus in the synchronized MRC5_VA cells released from the G1/S block treated with or without UV-C irradiation. The ZNFX1 gene locus is illustrated at the top. (D) RNA-seq expression of the ZNFX1 gene in the G1/S-phase-synchronized MRC5_VA cells during the DNA damage-induced cell cycle progression.
FigureS6.jpg
Extended Data 6 Figure S6. Molecular features of ZFAS1, Related to Figure 3 (A) The evolutionary conservation of ZFAS1 lncRNA over its entire length is depicted using the UCSC genome browser. (B) Schematic diagram showing sequences of identified 5’ and 3’ ends of ZFAS1 with RACE. (C) Sequences identified by RACE for the two variants (long and short isoforms) of ZFAS1. (D) Expression profiles of ZFAS1across various human tissues, adapted from NONCODEV6.
FigureS7.jpg
Extended Data 7 Figure S7. Coding capacity and expression dynamics of ZFAS1 in response to DNA damage, Related to Figure 3 (A) Protein-coding potential of ZFAS1predicted by PhyloCSF database, ACTB was presented as a positive control. (B) Agarose gel electrophoresis of eukaryotic expression recombination plasmids of pCMV6-Entry-ZFAS1-FL, pCMV6-Entry-ZFAS1-pORF, and pCMV6-Entry-ACTB by double digestion of Sgfl and Mlul. The marker was DL5000 at lane 1. (C) Cytoplasmic and chromatin extracts from MRC5_VA cells before and after UV-C treatment aligned to the human genome (hg38) and presented as coverage (normalized reads per million) as a function of chromosomal coordinate at the ZFAS1 loci. (D) qRT-PCR of ZFAS1 in the unsynchronized MRC5_VA cells treated with the indicated doses of different DNA damaging agents. Error bars represent ± SD from three independent experiments. Student’s t-Test, *p < 0.05. (E) qRT-PCR of ZFAS1 expression at the indicated time intervals after the DNA damage agent’s treatment in MRC5_VA cells. Error bars represent ± SD from three independent experiments. Student’s t-Test, *p < 0.05.
FigureS8.jpg
Extended Data 8 Figure S8. Epigenetic changes of ZFAS1 during DNA damage response, Related to Figure 3 (A) The chromatin accessibility in the regulatory regions of ZFAS1. The average ATAC-seq signal for the untreated and UV-irradiated MRC5_VA cells (at 3 h post-irradiation) was shown. A schematic representation of ZFAS1 locus was diagramed at the top. (B) DNA 5^mC methylation profiles of ZFAS1 around the TES (transcription ending site) as determined by RRBS-sequencing. Two biological replicates for each condition were shown.
FigureS9.jpg
Extended Data 9 Figure S9. ZFAS1 overexpression enhances cell viability after DNA damage, Related to Figure 4 (A) qRT-PCR analysis of ZFAS1 expression in MRC5_VA cells treated with three independent shRNA expression lentiviral plasmids targeting distinct regions of ZFAS1, showing averaged GAPDH-normalized data, relative to control cells (shCtrl). Each data point is presented as the means ± SD, n = 3. Student’s t-Test, *p < 0.05, **p < 0.01. (B) left: representative images of clonogenic survival assays of MRC5_VA cells overexpressing ZFAS1 under the indicated UV-C doses. right: percentage of surviving cells (logarithmic scale) plotted against UV-C dose. Error bars indicate the standard error of the mean from three independent experiments. Each data point is presented as the means ± SD, n = 3. *p < 0.05 compared to the WT cells (Student’s t-Test). (C) Short-term growth curve of MRC5_VA cells overexpressing ZFAS1 exposed to 10 J/m² UV-C irradiation. Untreated cells were used as controls. Each data point is presented as the means ± SD, n = 3. *p < 0.05 compared to the WT cells (Student’s t-Test). (D) Representative images of cells showing repair of CPDs at the indicated recovery time post UV irradiation using a specific antibody to CPDs (green signal) for the wild-type and MRC5 cells overexpressing ZFAS1. DAPI-stained nuclei in blue. Scale bar = 50 μm. (E) left: representative images from comet assay of cells treated with 10 J/m² UV-C and untreated controls. Scale bar = 100 μm. right: data from comet assay presented as box-whisker plots of the tail moment (A.U arbitrary units) of > 210 individual nuclei. ***p < 0.0001 (Student’s t-Test). The horizontal black bars represent the median tail moment for each dataset. (F) Bar graph representation of the flow cytometry profiles at the indicated time points after release from the G1/S block of control (wild-type) cells and ZFAS1-depleted cells without genotoxic stress. Each data point is presented as the means ± SD, n = 3. *p < 0.05 compared to wild-type cells (Student’s t-Test). (G) top: representative images of the EdU incorporation for the wide-type and ZFAS1-depleted MRC5_VA cells treated with UV-C irradiation (10 J/m²). Scale bar = 50 μm. White arrows indicate non-S-phase cells with unscheduled DNA synthesis (UDS) occurring. bottom: UDS levels were normalized and expressed as percentages of the cells with UDS in the total cell number in the bar chart. Error bars represent means ± SD. Student’s t-Test; n = 200 cells from three independent biological experiments.
FigureS10.jpg
Extended Data 10 Figure S10. DNA damage-induced exchange of RNAPIIo isoforms is dependent on ZFAS1, Related to Figure 5 (A) Representative images of the wild-type and ZFAS1-depleted MRC5_VA cells 3 h and 24 h after UV irradiation. Nascent EU-labeled RNA transcripts shown in green and DAPI-stained nuclei in blue. Scale bar = 100 μm. (B) Nascent mRNA production in different regions of the human KIFAP3 gene after release from DRB-inhibition. (C) Heatmaps illustrating the distribution of RNAPII-Ser5P reads without UV-C irradiation, as aligned at individual (rows) promoter-proximal regions (-250 bp to + 2 kb relative to TSSs) for WT and ZFAS1-depleted cells. (D) top: distributions of RNAPII-Ser5P read densities around TSSs and flanking regions (-250 bp to + 2kb relative to TSSs) without UV-C irradiation. bottom: the same as in (C) but categorized by gene expression status (active and inactive), as defined in previous report(Lavigne et al., 2017). (E) Heatmaps illustrating the Log₂ fold change (FC) of RNAPII-Ser5P comparing reads density between UV irradiation (UV+) at the indicated time points and steady-state (UV-) (FC = read density UV+/read density UV-), as aligned at individual regions (-250 bp to +2 kb relative to TSSs). (F) Same as in (E), but showing the Log₂ FC of RNAPII-Ser2P isoform.
FigureS11.jpg
Extended Data 11 Figure S11. ZFAS1 is not required for global transcription of mature mRNAs after UV stress, Related to Figure 6 (A) IGV snapshot showing tracks for ZFAS1 and its nearby coding gene ZNFX1 from RNA-seq of the wild-type and ZFAS1-depleted MRC5_VA cells treated with 10 J/m² UV-C irradiation. The ZFAS1 and ZNFX1 loci were diagramed on the top. (B) Principal component analysis (PCA) in the wild-type and two independent ZFAS1-depleted cell lines (KD1, KD2) upon UV-C irradiation from RNA-seq data. (C) RNA-seq analysis of differentially expressed genes (DEGs) in response to 10 J/m² UV-C irradiation in the wild-type and two independent ZFAS1-depleted cell lines. DEGs were considered at P value < 0.05 and Log₂ Fold change > 1. (D) Venn diagram showing the number of ATF3-dependent genes in the wild-type and ZFAS1-depleted MRC5_VA cells that were downregulated 3 hr upon UV-C irradiation. (E) qPCR analysis showing the activation and degradation of ATF3 mRNA in response to UV-C exposure in the WT and ZFAS1-depleted cells. Each data bar is presented as the means ± SD from three independent experiments. *p < 0.05 compared to undisturbed cells (Student’s t-Test).
FigureS12.jpg
Extended Data 12 Figure S12. Genome occupancy of ZFAS1 analyzed by ChIRP-seq in MRC5_VA cells prior to and after DNA damage induction, Related to Figure 6 (A) left: agarose gel electrophoresis showing that chromatin is sheared into the size range of 100-500 bp by sonication. right: qRT-PCR of RNA retrieval with the indicated ChIRP probe pools (LacZ as a negative control) from the wild-type MRC5_VA cells. (B) Tracks of the ZFAS1binding profiles at the ZFAS1 locus and its nearby gene loci from the ChIRP-seq data. The genomic coordinates of each gene are presented on the top. (C) Genome browser screenshots representing ZFAS1 ChIRP-seq reads count before (No UV) and after damage induction (UV-3 h) at the indicated genomic loci. TSS sites are illustrated as short green bars. (D) Average ZFAS1 ChIRP-seq profiles across all genes on the genome (hg38), before and after UV-C irradiation as indicated.
FigureS13.jpg
Extended Data 13 Figure S13. Increased genomic occupancy of ZFAS1 after UV stress, Related to Figure 6 (A) Heatmap of ZFAS1 ChIRP-seq signal in overlapping peak regions (raw peaks). Each row is a 4 kb genomic window centered on a ZFAS1 ChIRP peak in control and UV-irradiated cells; the peaks are aligned for all the ZFAS1 bound sites identified. Color bar indicates the number of ChIRP-seq reads. (B) Violin plots showing genome-wide distribution of ZFAS1 ChIRP-seq raw peak (overlapping peaks) signal in the control and UV-irradiated MRC5_VA cells lines. Two-sided Wilcoxon rank test (p < 2.2e-16) shows significant increase in chromatin occupancy for ZFAS1in response to UV-C treatment. (C) Distribution of narrow and broad peaks with distinct peak width. (D) Representative images of IGV screenshot of the ZFAS1 ChIRP-seq read densities and peaks called by MACS in the UV-irradiated cells. The genomic coordinates of each gene are presented on the top. (E) Proportion of ZFAS1-gain genes in intergenic, intragenic, and promoter regions of all ZFAS1-occupancy sites upon UV irradiation.
FigureS14.jpg
Extended Data 14 Figure S14. ZFAS1 mediates the recruitment of RNAPIIo isoforms on its target genes, Related to Figure 6 (A) ZFAS1 occupancy at the SMARCA2 loci from the ZFAS1 ChIRP-seq data. (B) RNAPIIo (top: CTD-ser5P; bottom: CTD-ser2P) distribution changes around the SMARCA2 promoter region after UV-C irradiation in the wild-type and the ZFAS1-depleted cell lines. (C) Same as in (A), but the XPC gene loci is presented. (D) Same as in (B), but the XPC gene promoter region is shown. (E) ZFAS1 binding sites upon UV-C irradiation enriched for several sequence motifs that are nearly identical to the four transcription factor motifs (KLF5, p53, PAX7, and RarA). (F) Percentage of interaction sites containing both ZFAS1 and RNAPIIo. Red bars, the proportion of the ZFAS1-bound genes with downregulated CTD-ser5P or upregulated CTD-ser2P within 6 hrs after UV-C irradiation in the wild type cell line of all genes with changed RNAPIIo upon UV-C. Blue bars, the proportion of the ZFAS1-bound genes containing either downregulated CTD-ser5P or upregulated CTD-ser2P within 6 hrs after UV-C exposure of all the ZFAS1-bound genes. White bars, the proportion of ZFAS1-bound genes with changed RNAPIIo only detected in the wild type cell line but not in the ZFAS1-depleted cell line of all the ZFAS1-bound genes with changed RNAPIIo. (G) Venn diagram showing the overlap between the ZFAS1-binding sites and the CPD hyper-hotspots after UV-C irradiation.
FigureS15.jpg
Extended Data 15 Figure S15. Identification of UV-induced ZFAS1 interactors, Related to Figure 6 (A) Venn diagram showing the number of the ZFAS1 interactors, identified by ChIRP-MS in the presence or absence of UV-C irradiation in MRC5_VA cells. (B) Top six GO-terms for the ZFAS1interactors. BP, biological process; CC, cellular component. (C) UV-induced ZFAS1 interactome. For clarity, only the interesting proteins are indicated, all the other interactors are labeled in black. (D) qPCR detection of ZFAS1retrieved by a NME-specific antibody in the RIP assay in MRC5_VA cells with or without UV-C irradiation. Each data bar is presented as the means ± SD from six independent experiments. ***p< 0.001 compared to undisturbed cells (Student’s t-Test).
FigureS16.jpg
Extended Data 16 Figure S16. In vivo characterization of Zfas1 genotypes, Related to Figure 7 (A) The expression levels of mouse lncRNA Zfas1 in various mouse tissues during embryonic development or postnatal stages. All expression data are normalized to mouse 18S RNA and presented as means ± SD from three independent experiments. (B) Expression of Zfas1 and Znfx1 genes in MEFs of indicated genotype, measured by RT-qPCR. Data are normalized to mouse 18S RNA and shown as the means ± SD, n = 3 for each genotype. **p < 0.05 (Student’s t-Test). (C) 8-week-old Zfas1 KO mice exhibit similar body weight compared to the littermate WT controls. Data are presented as means ± SD were analyzed with the Student’s t-Test. Male and female, n = 6:6. (D) Representative images of kidneys from 8-week-old female mice at experimental end point. Scale bar = 1 cm. (E) left, representative haematoxylin and eosin-stained (HE) sections of kidneys of mice. G indicates glomeruli. Typical examples of normal and abnormal glomeruli are shown. Scale bar = 50 μm. right, on one kidney section per animal of the indicated genotype the percentage of abnormal glomeruli was determined. Data are presented as means ± SD and were analyzed with the Student’s t-Test; *p < 0.05, **p < 0.01. male and female, n = 3:3. (F) Open filed test. Representative trajectories (top) and quantification of the total distance travelled and the velocity (bottom) for the WT and Zfas1-KO mice. Results are presented as means ± SD and were analyzed with the Student’s t-Test; *p < 0.05. Male and female, n = 6:6. (G) Morris water maze test. Escape latency to reach the hidden platform during the navigation test was recorded for the WT and Zfas1-KO mice. Data are presented as means ± SD and were analyzed with the student’s t-Test; *p < 0.05. Sample sizes for each group: n = 6. (H) Total distance, time spent in target quadrant, and number of platform crossings were shown in the probe test. Data are presented as means ± SD and were analyzed with the Student’s t-Test. Sample sizes for each group: n = 6. (I) Representative images of immunohistochemistry for GFAP performed on Zfas1^-/- and WT cerebral cortex sections, n = 3. Scale bar = 100 μm. (J) Representative images of the EdU incorporation for the WT and Zfas1^-/- mouse fibroblasts treated with UV-C irradiation (10 J/m²). Scale bar = 50 μm. White arrows indicate non-S-phase cells with unscheduled DNA synthesis (UDS) occurring.

Download PDF

Version 1

posted

You are reading this latest preprint version

lncRNA-ZFAS1, an emerging gate-keeper in DNA damage-dependent transcriptional regulation

Status:

Version 1

Abstract

Main

Results

Orchestration of gene expression dynamics during DNA damage response

Extensive and regulated lncRNA transcription of human fibroblasts during DNA damage-induced cell cycle progression

Discussion

Transcription regulation after DNA damage: difference in lncRNAs versus mRNAs

The importance of proper transcription shutdown and recovery on TC-NER

Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1