Alkyladenine DNA glycosylase associates with transcription elongation to coordinate DNA repair with gene expression

Abstract

Base excision repair (BER) initiated by alkyladenine DNA glycosylase (AAG) is essential for removal of aberrantly methylated DNA bases. Genome instability and accumulation of aberrant bases accompany multiple diseases, including cancer and neurological disorders. While BER is well studied on naked DNA, it remains unclear how BER efficiently operates on chromatin. Here, we show that AAG binds to chromatin and forms complex with RNA polymerase (pol) II. This occurs through direct interaction with Elongator and results in transcriptional co-regulation. Importantly, at co-regulated genes, aberrantly methylated bases accumulate towards the 3'end in regions enriched for BER enzymes AAG and APE1, Elongator and active RNA pol II. Active transcription and functional Elongator are further crucial to ensure efficient BER, by promoting AAG and APE1 chromatin recruitment. Our findings provide insights into genome stability maintenance in actively transcribing chromatin and reveal roles of aberrantly methylated bases in regulation of gene expression.

Fig. 1

AAG associates with active transcription and regulates gene expression. a Immunoblot analysis of chromatin fractionation assay indicating AAG distribution in total fraction (TF), soluble supernatant fraction (SF) and chromatin fraction (CF). Histone H3 and α-tubulin served as controls. b Quantification of three independent experiments as the one depicted in a; error bars represent mean ± SEM (n  = 3). c Immunoprecipitation of AAG from HEK293T whole-cell extracts (WCEs) untreated or treated with DNaseI, Mnase, and RNaseI, showing the interaction with RNA polymerase II phosphorylated at Serine 2 (S2P) of CTD repeat. d Immunoblot analysis of WCEs from HEK293T WT and AAG^−/− cells generated by CRISPR-Cas9 technology. e Heat map of the expression levels of AAG-regulated genes in HEK293T cells. Color scale is representing change in the gene expression depicting log2 fold change relative to the mean. f Top six biological processes (BP) gene ontology (GO) terms as determined by the Database for Annotation, Visualization and Integrated Discovery (DAVID) for genes dysregulated in HEK293T AAG^−/− cells when compared to WT. g Up- and down-regulated differentially expressed genes (DEGs) in HEK293T AAG^−/− cells. h, i Top six BP GO terms as determined by DAVID for genes upregulated (h) and downregulated (i) genes. Source data are provided as Source Data file.

Fig. 2

AAG directly interacts with ELP1 subunit of transcriptional Elongator complex. a Silver staining analysis of untreated or MMS (methyl methanesulfonate) treated FLAG-AAG immunoprecipitation (IP) samples subjected to LC/MS-MS analysis. IP sample from HEK293T cells transfected with empty FLAG vector served as negative control. LC/MS-MS analysis of FLAG-AAG IP samples identified ELP1 as the most specifically enriched AAG interacting partner. Remaining subunits of core Elongator (ELP2 and 3), and known interacting partner PCNA are depicted. See also Supplementary Data 2 for the complete list of proteins identified in all samples. b IP of AAG from HEK293T whole cell extracts (WCEs) treated or untreated with MMS. c Immunoblot analysis of gel shift experiments with purified recombinant AAG and FLAG-ELP1 proteins, and chemical crosslinking with indicated amounts of BS³ (bissulfosuccinimidyl suberate). Control samples – proteins in the absence of crosslinking agent (lanes 1-3); and single proteins with highest amount of BS³ (lanes 6, 7). d Schematic representation of full-length AAG, AAG lacking 80 N-terminal amino acids (∆80 AAG), and first 80 N-terminal amino acids of AAG (1-80aa AAG); numbers indicate amino acids. e Co-IP of purified FLAG-ELP1 with AAG full length, or AAG lacking 80 N-terminal amino acids (∆80 AAG). f IP of GFP-tagged first 80 N-terminal amino acids of AAG (1-80aa AAG) (GFP-1-80aa AAG) expressed in HEK293T AAG^−/− cells. GFP-tag positioned N-terminally. g Proximity ligation assay (PLA) showing the AAG-ELP1 interaction in HEK293T cells. Scale bar: 50 μm. Source data are provided as a Source Data file.

Fig. 3

Elongator and AAG co-regulate gene expression. a Immunoblot of HEK293T whole-cell extracts (WCEs) from WT, AAG^−/−, ELP1^−/−, and AAG^−/−ELP1^−/− cell lines generated via CRISPR-Cas9 technology. b Top six biological processes (BP) gene ontology (GO) terms as determined by DAVID for genes dysregulated in HEK293T ELP1^−/− cells when compared to WT cells. c Downregulated and upregulated differentially expressed genes (DEGs) in HEK293T ELP1^−/− cells. d Venn diagrams of AAG- and ELP1-regulated genes in HEK293T cells. e Top six BP GO terms as determined by DAVID of genes co-regulated by AAG and ELP1 in HEK293T cells. f Directionality of DEGs regulated by AAG and ELP1 in HEK293T cells. g–j Relative mRNA levels of ALDH1A2 (g), CRMP1 (h), CDH23 (i) and YTHDC1 (negative control) (j) genes in HEK293T WT, AAG^−/−, ELP1^−/− and AAG^−/−ELP1^−/− cells. Error bars represent mean ± SEM (n ≥ 3). **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001, one-way ANOVA. Source data are provided as Source Data file.

Fig. 4

Elongator, components of AAG-initiated BER and AAG substrates accumulate towards the 3′end of co-regulated genes. a AAG ChIP-qPCR assays in HEK293T WT cells comparing percentage of input at gene bodies of unaffected gene (YTHDC1) and differentially expressed genes (ALDH1A2, CRMP1, CDH23, SYT9, and CDH4). b–e ChIP-qPCR assays showing relative AAG occupancy in AAG- and ELP1- dependent genes ALDH1A2 (b), CDH23 (c), CRMP1 (d), and unaffected gene YTHDC1 (e) in HEK293T WT cells. f–h ChIP-qPCR assays showing relative occupancy of HA-ELP1 in AAG- and ELP1- dependent ALDH1A2 (f), CDH23 (g), and CRMP1 (h) genes in HEK293T HA-ELP1cells. i–l ChIP-qPCR assays showing relative APE1 occupancy in negative control YTHDC1 gene (i), and AAG- and ELP1-dependent ALDH1A2 (j), CDH23 (k), CRMP1 (l) genes in HEK293T WT cells. m–p qPCR DNA damage assay showing differences in distribution pattern of aberrant AAG substrates in unaffected gene YTHDC1 (m) and genes regulated by AAG and ELP1: ALDH1A2 (n), CDH23 (o), CRMP1 (p) in HEK293T WT cells. Values are shown as relative occupancy: % input of specific gene region, relative to percentage input of promoter region. Error bars indicate mean ± SEM (n ≥ 3). Two-tailed Student’s t-test in a; one-way ANOVA in b–p; *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. Source data are provided as Source Data file.

Fig. 5

Lack of functional Elongator hampers AAG and APE1 chromatin recruitment, and AAG-initiated repair. a–d ChIP-qPCR experiments comparing AAG occupancy at promoters and gene bodies of ALDH1A2 (a), CRMP1 (b), CDH23 (c), and YTHDC1 (d) in HEK293T WT and ELP1^−/− cells. e Immunoblot analysis of AAG levels in chromatin fraction of HEK293T WT and ELP1^−/− cells. Histone H3 served as control. f Quantification of experiments as the one depicted in e. g ChIP-qPCR experiments comparing APE1 occupancy at gene bodies of YTHDC1, ALDH1A2, CRMP1, CDH23, SYT9, and CDH4 in HEK293T WT and ELP1^−/− cells. Error bars represent the SEM calculated from at least three independent experiments. h Immunoprecipitation of AAG from HEK293T WT and ELP1^−/− whole cell extracts showing the interaction with RNA polymerase II phosphorylated at Serine 2 (S2P) of CTD repeat. i Measurement of AAG DNA glycosylase activity in HEK293T WT, AAG^−/− and ELP1^−/− on hypoxanthine (Hx)-containing plasmid by FM-HCR assays. j AAG DNA glycosylase activity determined by FM-HCR assays on hypoxanthine (Hx)-containing plasmid in HEK293T WT cells complemented with GFP and AAG^−/− cells complemented with GFP, GFP-AAG or GFP-Δ80 AAG. Error bars represent mean ± SEM (n ≥ 3). Two-tailed Student’s t-test (a–d, f, g); one-way ANOVA (i, j), NS – non significant, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. Source data are provided as Source Data file.

Fig. 6

Transcription inhibition reduces AAG occupancy and impairs removal of AAG substrates. a–d ChIP-qPCR experiments comparing AAG occupancy in DMSO and DRB treated HEK293T WT cells at ALDH1A2 (a), CRMP1 (b), CDH23 (c), and YTHDC1 (d) genes. Error bars, SEM from three independent experiments. e AAG Comet-FLARE assay in DMSO or DRB treated HEK293T cells. Error bars indicate mean ± SEM (n ≥ 3). **p ≤ 0.01; ***p ≤ 0.001, one-way ANOVA. Source data are provided as Source Data file.

Fig. 7

Model of AAG-initiated DNA repair coordinated with gene expression. Elongator complex associates with RNA pol II to promote transcription elongation, accumulating towards 3′end of regulated genes. AAG through its unstructured N-terminal region associates with ELP1 subunit of Elongator, thus forming complex with active transcription machinery. During active transcription chromatin is locally decondensed, which allows AAG to efficiently initiate BER by recognizing and removing aberrant bases. AAG-initiated BER likely temporarily inhibits RNA pol II progression, thus resulting in reduced expression of co-regulated genes. In the absence of Elongator, transcription of target genes is repressed, while AAG recruitment to chromatin and initiation of BER is impaired. For more details see text. Schematic representation was created with Biorender.com.