Fig 2: The TFIIH holoenzyme suppresses telomere fragility and replication defects. (A) Cartoon of the TFIIH holoenzyme subunits. (B) Immunoblot to monitor XPB levels after CRISPR/Cas9 targeting (96 h) of the indicated subunits of TFIIH. Loss of XPB confirms the knockout of the targeted genes. γ-Tubulin served as the loading control. (C) Quantification of long arm fragile telomeres detected by FISH in Trf1F/F MEFs with CRISPR/Cas9 targeting of the subunits of the TFIIH core as indicated. (D) Immunoblot to monitor the level of MNAT1, CDK7, CCNH, and XPA after dual-cutting CRISPR/Cas9 targeting of the corresponding genes (96 h). γ-Tubulin served as the loading control. (E) Quantification of long arm fragile telomeres detected by FISH in Trf1F/F MEFs with CRISPR/Cas9 targeting of the indicated subunits of the CAK complex or the NER proteins as in D. (F) Immunoblot to monitor the effect of 60 μM SP treatment (25 h) on XPB. γ-Tubulin served as the loading control. (G) Quantification of long arm fragile telomeres detected by FISH in Trf1F/F MEFs treated with 60 μM SP or 0.25 μM CDK7i for 25 h. (H) Quantification of telomere molecules with IdU signals by SMARD in Trf1F/F MEFs with CRISPR/Cas9 targeting of Luc, Xpb, or Xpd (96 h). Data are means ± SD from three experiments, with ∼100 telomere molecules analyzed per experiment. (I) MiDAS assay to detect CFS expression. MiDAS was detected in cells treated with 10 μM EdU 1 h prior to harvesting of metaphases. Trf1F/F MEFs were untreated or treated with 60 μM SP, 0.25 μM CDK7i, or 0.4 μM aphidicolin for 24 h prior to EdU labeling. EdU was detected by Click chemistry (green). DNA was stained with DAPI (red). (J) Quantification of the number of EdU foci per chromosome in Trf1F/F MEFs treated with the corresponding drugs as in I. Data are the means ± SD of three experiments, each with 20 metaphases. (K) Quantification of nontelomeric 53BP1 foci in Trf1F/F MEFs with CRISPR/Cas9 targeting of Luc, Xpb, or Xpd (96 h). Data are means ± SD of three independent experiments of >50 nuclei each. (L) Immunoblot to monitor the level of CHK1 and phospho-CHK1 (pCHK1 S345) after CRISPR/Cas9 targeting of Luc, Xpb, or Mnat1 with cells treated with 5 μM aphidicolin as the positive control. GAPDH served as the loading control. All metaphase analyses are means ± SD of three or more experiments of ∼2000 telomeres analyzed. P-values were derived from two-tailed unpaired t-test. (****) P ≤ 0.0001, (***) P ≤ 0.001, (**) P ≤ 0.01, (*) P ≤ 0.05, (n.s.) P > 0.05.

Fig 3: Epistasis and interaction of TFIIH and TRF1. (A) Immunoblot to monitor XPB, XPD, and TRF1 in Trf1F/F MEFs ± Cre (96 h) with CRISPR/Cas9 targeting of Luc, Xpb, or Xpd (96 h) as indicated. γ-Tubulin served as the loading control. (B) Quantification of long arm fragile telomeres detected by FISH in Trf1F/F MEFs ± Cre (96 h) with CRISPR/Cas9 targeting of Luc, Xpb, or Xpd (96 h) as in A. (C) Anti-Myc co-IPs of Myc-tagged mouse TRF1 and FLAG-tagged mouse TFIIH subunits from cotransfected 293FT cells. Immunoblots were probed with anti-Myc and anti-FLAG antibodies. Four percent of the lysate was run as the input. (D) Anti-Myc co-IPs of Myc-tagged mouse TRF1, TRF1-E83K/LW-TI, and TRF2 and FLAG-tagged mouse XPB, GTF2H3, and GTF2H4 from cotransfected 293FT cells. IP samples were treated with 25 U of benzonase where indicated. Immunoblots were probed with anti-Myc and anti-FLAG antibodies. Four percent of the lysate was run as the input. (E) Immunoblot to monitor XPB and TRF1 in Trf1F/F MEFs with overexpression of wild-type TRF1 and TRF1-E83K/LW-TI and with CRISPR/Cas9 targeting of Luc or Xpb (96 h). γ-Tubulin served as the loading control. (F) Quantification of long arm fragile telomeres detected by FISH in Trf1F/F MEFs as in E. All metaphase analyses are means ± SD of three experiments of ∼2000 telomeres analyzed. P-values were derived from two-tailed unpaired t-test. (****) P ≤ 0.0001, (***) P ≤ 0.001, (**) P ≤ 0.01, (n.s.) P > 0.05. (G) Model of how different motifs in TRF1 function to suppress telomere fragility, replication defects, ATR signaling, and STAs.

Fig 4: XPB and XPD are required for the suppression of telomere fragility. (A) Immunoblot for XPB and XPD to monitor expression after dual-cutting CRISPR/Cas9 targeting of the two genes (96 h). γ-Tubulin served as the loading control. (B) Examples of fragile telomeres caused by targeting of Luc, Xpb, or Xpd (96 h) detected by FISH and CO-FISH as indicated. (C,D) Quantification of long arm fragile telomeres detected by FISH in C and CO-FISH in D in Trf1F/F MEFs as in B. (E) Quantification of 53BP1 TIFs in Trf1F/F MEFs as in B. Data are means ± SD of three independent experiments of >50 nuclei each. (F) Quantification of long arm STAs detected by FISH in Trf1F/F MEFs as in B; Trf1F/F MEFs + Cre served as a positive control. All metaphase analyses are means ± SD of three experiments of ∼2000 telomeres analyzed. All P-values were derived from two-tailed unpaired t-test. (***) P ≤ 0.001, (**) P ≤ 0.01, (*) P ≤ 0.05, (n.s.) P > 0.05.

Fig 5: Artificial antisense oligonucleotides (ASOs) designed to induce GTF2H4 alternative splicing facilitate p52 C-terminal truncation and promote TFIIH stability.(A) Schematic representation of the GTF2H4 gene structure and the design of ASOs to induce intron 13 inclusion, leading to a frameshift and a C-terminal truncation of p52. In-frame stop codons in intron 13 (orange). (B) Induction of p52 C-terminal truncation in TTD-A cells treated with ASOs (24 hours). (C and D) Simultaneous blocking of the 5’ and 3’ splice sites of intron 13 facilitates efficient induction of the p52 truncation and TFIIH stabilization in TTD-A cells. In D, immunoprecipitation of XPB/p89 was performed. Asterisks indicate nonspecific products. (E) ASO treatment mostly rescues transcription in TTD cells. Transcription activity was measured by EU incorporation. Bars and error bars represent means and SEM, respectively, of experiments (n = 3, as indicated by the colored-circles and their corresponding plots). 20J/m2 UVC-irradiation. wo, without treatment; control, control-ASO treatment.