Identification of modulators of the ALT pathway through a native FISH-based optical screen

Abstract

A significant portion of human cancers utilize a recombination-based pathway, alternative lengthening of telomeres (ALT), to extend telomeres. To gain further insights into this pathway, we developed a high-throughput imaging-based screen named TAILS (telomeric ALT in situ localization screen) to identify genes that either promote or inhibit ALT activity. Screening over 1,000 genes implicated in DNA transactions, TAILS reveals both well-established and putative ALT modulators. Here, we present the validation of factors that promote ALT, such as the nucleosome-remodeling factor CHD4 and the chromatin reader SGF29, as well as factors that suppress ALT, including the RNA helicases DExD-box helicase 39A/B (DDX39A/B), the replication factor TIMELESS, and components of the chromatin assembly factor CAF1. Our data indicate that defects in histone deposition significantly contribute to ALT-associated phenotypes. Based on these findings, we demonstrate that pharmacological treatments can be employed to either exacerbate or suppress ALT-associated phenotypes.

Keywords: ALT; CP: Molecular biology; DDX39A; histone deposition; ssTelo; telomeres.

Figure 1.. TAILS identifies modulators of ALT

(A) Rolling circle assay (RCA) analysis of genomic DNA isolated from U2OS cells transfected with the sgRNA and relative quantification. (B) Example of TAILS data and quantification; U2OS cells transfected with indicated sgRNA were processed as described in the TAILS pipeline. A minimum of 2,200 cells were analyzed per condition. The scale bar represents 10 μm. (C) TAILS analysis in ALT-positive U2OS cells. The ssTelo intensity derived from two biological replicates of the arrayed sgRNA library. Linear regression analysis by Pearson correlation coefficient (R) was calculated, and it is 0.701. (D) Scatterplot displaying mean Z score for each gene assayed (y axis) and the relative ranking based on descending Z score (x axis). Dotted lines indicate the cutoff value (±2.4) chosen to identify putative hits. Genes that have previously been reported to affect ALT activity are indicated. (E) Images obtained by TAILS of hits selected known modulators of ALT. The scale bar represents 10 μm. (F) Gene function classification of all the genes contained in sgRNA library (left diagram) and the hits identified by TAILS (right diagrams). For more information, see Table S1.

Figure 2.. SGF29, CHD4, and SUMOylation are required for ALT activity

(A) Scatterplot shows the Z score for genes identified by TAILS as putative ALT activators. (B) Gene function classification of the genes displayed in (A). (C and D) ssTelo staining and relative quantification of U2OS cells infected with shRNA against SGF29 (shSGF29–1, shSGF29–2, and shSGF29–3) or control (wild type [WT]). Each dot represents one cell, with a minimum of 300 cells per condition analyzed across 2 independent experiments. (E and F) ssTelo staining and relative quantification of U2OS cells transfected with siRNA against CHD4 (siCHD4–1 and siCHD4–2) or a non-targeting control (siCtrl). Each dot represents one cell, with a minimum of 500 cells per condition analyzed across 3 independent experiments. (G) U2OS cells transfected with siRNA against CHD4 (siCHD4–1) or a non-targeting siRNA (siCtrl) were stained for PML (red) and TRF2 (green). (H) Quantification of data shown in (G) with graphs indicating the percentage of cells with at least 3 PML-TRF2 co-localizations (APBs) per nucleus, defined as two foci overlapping by 50% or more. Error bars represent the standard deviation of the mean, with a minimum of 700 cells per condition analyzed across 4 independent experiments. (I) U2OS cells co-transfected with vectors expressed CHD4-GFP (green) and either WT or catalytically dead (D480A) TRF1-FokI-mCherry (red). (J) Quantification of data shown in (I) with graphs indicating the percentage of cells with at least 3 CHD4-TRF1-FokI co-localizations per nucleus, defined as two foci overlapping by 50% or more. Error bars represent the standard deviation of the mean, with a minimum of 500 cells per condition analyzed across 3 independent experiments. (K) ssTelo staining of ALT-positive U2OS, SAOS-2, and G292 cells treated with SUMOi compared to untreated sample. (L) Quantification of ssTelo analysis shown in (K) and Figure S2H. Each dot represents one cell, with a minimum of 300 cells per condition analyzed across 2 independent experiments. The scale bar represents 10 μm. Data are represented as mean ± SEM. An unpaired t test was used for statistical analysis; *p ≤ 0.05, ***p ≤ 0.001, and ****p ≤ 0.0001 on the graphs.

Figure 3.. The RNA helicases DDX39A/B are ALT suppressors

(A) Gene function composition of the arrayed sgRNA library (“Library”) and the number genes that were identified as putative ALT suppressors by TAILS (“Enriched”). (B) List of ALT-suppressor genes that were further characterized in this study. The table reports the average Z score (“Score”) and gene function category (“Function”). (C and D) Rolling circle assay (RCA) analysis of genomic DNA isolated from 3 independent DDX39A^−/− clones (C1, C2, and C3) and the parental U2OS cells (WT). (E and F) Representative images and quantification of DDX39A-deficient and -proficient cells in ALT-positive U2OS and ALT-negative HeLa backgrounds. Each dot represents one cell, with a minimum of 200 cells per condition analyzed across 2 independent experiments. The scale bar represents 10 μm. (G) ssTelo staining of U2OS cells transfected with sgRNAs against FANCM, DDX39A, DDX39B, and a non-targeting control (sgCtrl) in the presence of absence of sgRNA against BLM. A minimum of 1,800 cells were analyzed per condition. The scale bar represents 10 μm. (H) Quantification of the data shown in (G). (I and J) Representative images and quantification of ssTelo analysis ALT-positive U2OS, G292, and SAOS-2 cells treated with transcription inhibitor DRB. Each dot represents one cell, with a minimum of 250 cells per condition analyzed across 2 independent experiments. The scale bar represents 10 μm. (K) Quantification of ssTelo analysis U2OS cells treated with DRB in the presence of absence of sgRNA against BLM. Each dot represents one cell, with a minimum of 300 cells per condition analyzed across 2 independent experiments. For representative images, see Figure S4C. Data are represented as mean ± SEM. An unpaired t test was used for statistical analysis; **p ≤ 0.01 and ****p ≤ 0.0001 on the graphs.

Figure 4.. Reduced histone deposition elevates ssDNA at ALT-positive telomeres

(A and B) Representative images and quantification of ssTelo staining U2OS cells treated with siRNAs against either CHAF1B or HIRA histone chaperones encoding genes. (C) Quantification of de novo histone deposition in U2OS cells treated with siRNAs against CHAF1B, HIRA, or a non-targeting control (siCtrl). Each dot represents one cell, with a minimum of 550 cells per condition analyzed across 2 independent experiments. TMR (tetramethylrhodamine) intensity levels are quantified and averaged, normalized to samples treated with siCtrl, and plotted on a log scale. (D and E) Representative images and quantification of ssTelo staining U2OS cells treated with siRNAs against TIMELESS (siTIM) compared to control cells (siCtrl). Each dot represents one cell, with a minimum of 250 cells per condition analyzed across 2 independent experiments. (F) Quantification of de novo histone deposition in U2OS cells treated with siRNAs against TIMELESS. Each dot represents one cell, with a minimum of 900 cells per condition analyzed across 2 independent experiments. The scale bar represents 10 μm. Data are represented as mean ± SEM. An unpaired t test was used for statistical analysis; ns p > 0.05, **p ≤ 0.01, and ****p ≤ 0.0001 on the graphs.

Figure 5.. Inhibition of ATR or TLK elevates telomeric ssDNA and reduces histone deposition

(A–C) Representative images and quantification of ssTelo staining of a panel of ALT-positive (U2OS, LM216J, G292, and SAOS-2) and ALT-negative (HeLa and LM216T) cells treated with TLKi or ATRi for 6 h. Each dot represents one cell, with a minimum of 250 cells per condition analyzed across 2 independent experiments. (D and E) Representative images and quantifications of ssTelo staining of U2OS in the presence or absence of sgRNA against BLM treated with TLKi or ATRi. Each dot represents one cell, with a minimum of 250 cells per condition analyzed across 2 independent experiments. (F and G) Representative images of TMR (tetramethylrhodamine) staining (red) of H3.1-SNAP U2OS cells treated with TLKi (F), ATRi (G), or control (veh [vehicle]). (H) Quantification of de novo H3.1 histone deposition in U2OS cells treated with TLKi or ATRi. Each dot represents one cell, with a minimum of 800 cells per condition analyzed across 2 independent experiments. The scale bar represents 10 μm. Data are represented as mean ± SEM. An unpaired t test was used for statistical analysis; ns p > 0.05, *p ≤ 0.05, ***p ≤ 0.001, and ****p ≤ 0.0001 on the graphs.