Requirement of transcription-coupled nucleotide excision repair for the removal of a specific type of oxidatively induced DNA damage

Abstract Accumulation of DNA damage resulting from reactive oxygen species was proposed to cause neurological and degenerative disease in patients, deficient in nucleotide excision repair (NER) or its transcription-coupled subpathway (TC-NER). Here, we assessed the requirement of TC-NER for the repair of specific types of oxidatively generated DNA modifications. We incorporated synthetic 5′,8-cyclo-2′-deoxypurine nucleotides (cyclo-dA, cyclo-dG) and thymine glycol (Tg) into an EGFP reporter gene to measure transcription-blocking potentials of these modifications in human cells. Using null mutants, we further identified the relevant DNA repair components by a host cell reactivation approach. The results indicated that NTHL1-initiated base excision repair is by far the most efficient pathway for Tg. Moreover, Tg was efficiently bypassed during transcription, which effectively rules out TC-NER as an alternative repair mechanism. In a sharp contrast, both cyclopurine lesions robustly blocked transcription and were repaired by NER, wherein the specific TC-NER components CSB/ERCC6 and CSA/ERCC8 were as essential as XPA. Instead, repair of classical NER substrates, cyclobutane pyrimidine dimer and N-(deoxyguanosin-8-yl)-2-acetylaminofluorene, occurred even when TC-NER was disrupted. The strict requirement of TC-NER highlights cyclo-dA and cyclo-dG as candidate damage types, accountable for cytotoxic and degenerative responses in individuals affected by genetic defects in this pathway.


Supplementary Figures and Table pages 2 -10
Supplementary Figure S1. The synthesis procedure of the 2′F-Tg phosphoramidite.
Supplementary Figure S6. Generation of reporter constructs harbouring synthetic cyclopurine adducts (c-dA or c-dG) or thymine glycol (Tg or 2′F-Tg) in the 5′-untranslated region of the EGFP gene.
Supplementary Figure S8. Demonstration of transcription blockage by synthetic cis-syn cyclobutane TT dimer (T□T) in the transcribed DNA strand of the EGFP gene.
Supplementary Table S1. Positions and nucleotide sequences of sgRNA pairs used for CRISPR-Cas9 editing of the specified gene loci and the list of PCR primers used to screen single cell clones for the presence of specific deletions.
(B) CSA gene disruption scheme. A pair of sgRNAs (arrowheads) was used to delete the fragment including the WD 4-coding exons 6, 7, and 8 (bracket). SgRNA sequences are reported in Supplementary Table S1.
(C) Demonstration of deletion of the targeted CSA fragment in the 2C5 clone (bold typeface) chosen from 52 clones isolated by single cell sorting and screened by PCR. A primer pair flanking the upstream sgRNA binding site (Supplementary Table 1) was used to amplify the native (unedited) CSA allele.
(D) Validation of CSA knockout in the 2C5 clone by Western blotting with EPR9237 rabbit monoclonal antibody (Abcam, ab137033). Mouse monoclonal antibody to HSP90 (Santa Cruz Biotechnology #sc-13119) was used as an equal loading control.
(E) Validation of the TC-NER-deficient phenotype of the CSA knockout clone. Host cell reactivation assay was carried out using pZAJ-5C construct harboring synthetic dG(N 2 )-AAF adduct, whose repair strictly requires TC-NER (4). Parental HeLa cells (WT) were analysed in parallel as a TC-NERproficient control. Fluorescence distribution plots of transfected cells (representative experiment) and quantification of the recovered fraction of the EGFP expression (dG(N 2 )-AAF/dG) in duplicate samples transfected with the same constructs (mean ± SD). (B) DDB2 gene targeting using a pair of sgRNAs (arrowheads) flanking the region encompassing the exons 4, 5, and 6 (magnified in the cartoon). The exon 6 contains the entire WD 4 coding sequence (bracket). SgRNA sequences are reported in Supplementary Table S1.
(C) Demonstration of the DDB2 exon 6 deletion in two selected (of 30 screened) knockout clones (1D11 and 2C6) isolated by single cell sorting. Parental HeLa cells and the CSA knockout clone 2C5 (Supplementary Figure S2) were analyzed in parallel. One primer pair was designed to specifically amplify the rearranged (edited) DDB2 gene fragment; another pair flanking the upstream sgRNA target site was used to amplify the native (unedited) DDB2 fragment (Supplementary Table S1).
(D) Western blotting with the XPE/DDB2 rabbit polyclonal antibody (R&D Systems #AF3297). All cell extracts display a faint non-specific band slightly above the DDB2 level but there is no detectable DDB2 band in the knockout clones (1D11 and 2C6). Mouse monoclonal antibody to HSP90 (Santa Cruz Biotechnology #sc-13119) was used as an equal loading control. (A) Scheme of the NTHL1 gene targeting using a pair of sgRNAs homologous to exons 1 and 5 (arrowheads). The generated deletion includes almost the entire protein coding sequence, including the catalytic lysine (K155) codon within the exon 3. SgRNA sequences are reported in Supplementary  Table S1.
(B) PCR demonstration of the NTHL1 gene deletion in chosen clones, as compared to the parental HeLa cell line (WT). A primer pair enclosing both sgRNA sites was used to amplify the rearranged (edited) NTHL1 fragment; another pair was used to reject the clones displaying the native (unedited) fragment (Supplementary Table S1). Clones marked by bold typeface show only the band specific to the rearranged locus. The absence of the NTHL1 protein in the indicated clones was further confirmed by Western blotting (Figure 6).

Conditions for the synthesis of the 2′F-Tg phosphoramidite
All reagents and solvents were purchased from commercial sources and used without further purification, unless otherwise stated. All reactions were carried out under argon atmosphere. Spectrometry experiments were carried out in a Q-TOF Bruker MaXis Impact HR-Mass Spectrometer.
Selected 1 H, 13 C, COSY, HSQC and HMBC NMR spectra of compounds 2-9 are reported (Extended data 1-33). (2) To a solution of 1 (15.5 g, 50 mmol, 1 eq) in anhydrous DMF (50 mL) were added diphenyl carbonate (12.9 g, 60 mmol, 1.2 eq) and NaHCO3 (250 mg, 2.5 mmol, 0.05 eq) and the mixture was heated to reflux for 3 hours. The reaction mixture was cooled to room temperature and water was added. The water layer was washed 4 times with diethyl ether and then was evaporated under reduced pressure.

2,2′-Αnhydrothymidine
The residue was recrystallized by MeOH and dried under reduced pressure to give 9.80 g (82%) of 2 as white powder.
The organic layer was dried over Na2SO4, evaporated under reduced pressure, and purified by column

3′,5′-Ο-dibenzyl-2′-fluorothymidine glycol (5)
To a solution of OsO4 (700 mg, 2.6 mmol, 1 eq) in dry pyridine (3 mL) was added a solution of 4 (1.150 mg, 2.6 mmol, 1 eq) in dry pyridine (3.5 mL). The reaction mixture was stirred in room temperature for 2.5 h and then aqueous solution of NaHSO3 (1.8 g NaHSO3 in 1 mL H2O) and pyridine (7 mL) were added. The brown mixture was stirred for 12 hours at room temperature and then was extracted with DCM, washed with water, brine and the organic layer was dried over Na2SO4 and evaporated to give 1.20 g (98%) 5 as a mixture of diastereomers.
The products were characterized as following: