TRIM24 directs replicative stress responses to maintain ALT telomeres via chromatin signaling

Summary

An inability to replicate the genome can cause replication stress and genome instability. Here, we develop BLOCK-ID in human cancer cells, a proteomic method to identify and visualize proteins at stressed replication forks. This approach identified mediators of the replication stress response, including the chromatin acetylation reader protein TRIM24. We uncovered a crucial role for TRIM24 in coordinating Alternative Lengthening of Telomeres (ALT), a replication stress-directed telomere extension mechanism. Our data reveal that TRIM24 is recruited to telomeres via a p300/CBP-dependent acetylation chromatin signaling cascade to organize the assembly of ALT-associated PML bodies (APBs) and promote de novo telomere DNA synthesis. Tethering of TRIM24 at telomeres was sufficient to stimulate de novo telomere DNA synthesis in a SUMO-dependent, but p300/CBP- and PML-independent manner. Collectively, these findings uncover an indispensable epigenetic signaling pathway involving TRIM24 and p300/CBP that mediate ALT telomere maintenance.

eTOC

Kim et al. develop BLOCK-ID, an approach to capture replication stress response factors. The chromatin reader TRIM24 was identified as a critical mediator of Alternative Lengthening of Telomeres (ALT) through a mechanism involving histone acetylation and SUMOylation signaling that when disrupted results in aberrant telomere function in ALT.

Graphical Abstract:

Introduction

Faithful transmission of genetic and epigenetic material in eukaryotes requires the assembly of replicated DNA into chromatin¹. Histones, the building blocks of chromatin, are highly modified by post-translational modifications (PTMs) that govern many cellular functions, including during DNA replication. Replication stress is an oncogenic driver causing mutagenesis, chromosomal rearrangements, and genome instability^2,3. Proteomic profiling of stressed replication forks has revealed networks of DNA damage response (DDR) factors that respond to insults during DNA replication. These approaches include iPOND^4–6 and Nascent Chromatin Capture (NCC)⁷ have revealed a systemic reorganization of chromatin at stalled and damaged replication forks. Other studies have shown the assembly of heterochromatin⁸ and chromatin remodeling complexes to protect against transcription-replication conflicts⁹. Despite these advancements in proteomic methodologies that primarily rely on nucleotide analog containing nascent DNAs^4,6,7, an incomplete inventory of the stressed replication-specific responders, particularly epigenetic regulators, limits our understanding of mechanisms that respond to replication stress.

Here, we developed BLOCK-ID to overcome these limitations. BLOCK-ID couples the LacO/LacI replication barrier system and proximity-dependent biotin identification method with proteomics and fluorescence microscopy to identify spatially associated DDR and chromatin factors at blocked replication forks. We uncovered proteins enriched at blocked replication forks with BLOCK-ID, including the bromodomain (BRD) chromatin reader protein, Tripartite Motif containing 24, TRIM24. Our functional analyses uncovered an unexpected and crucial role for TRIM24 in mediating Alternative Lengthening of Telomeres (ALT), a pathological telomere extension mechanism detected in advanced cancer subtypes linked with mutations in chromatin modifiers, including ATRX-DAXX chromatin remodeling complex and histone H3.3^10,11. The ensuing alterations in chromatin dynamics cause pronounced replicative stress that stimulates homology-directed repair (HDR) mechanisms, like break-induced replication (BIR), which repairs and lengthens telomeres in those so-called ALT cancer cells^12–14. These activities that extend telomeres have been visualized within unique nuclear compartments termed ALT-associated PML Bodies (APBs)¹⁵. APB formation involves multivalent contacts between SUMOylated proteins and SUMO-interaction motifs (SIM) present within PML and mediators of ALT-HDR, such as RAD52 and BLM^16–18. It is proposed that this organization provide an optimized environment for telomere extension as it sequesters telomeres within a DDR-factor-rich sub-nuclear compartment. Indeed, de novo telomere DNA synthesis is predominantly observed within APBs¹⁹ and disrupting PML and key mediators like BLM attenuate the ALT mechanism, leading to gradual telomere attrition²⁰.

This study demonstrates that TRIM24 fulfills crucial roles in maintaining functional telomeres by the ALT mechanism. We present evidence that TRIM24-deficient ALT cancer cells exhibit an inefficient and perturbed ALT mechanism, most notably impaired APB formation and de novo telomere DNA synthesis, robust telomere shortening, and compromised telomere integrity due to elevated replicative stress. Furthermore, this study reveals a PTM-driven epigenetic signaling pathway involving p300/CBP-dependent histone acetylation and PIAS1-dependent SUMOylation that coordinates TRIM24’s critical functions in maintaining a productive ALT mechanism. Remarkably, we found that artificially tethering TRIM24 at telomeres can bypass the requirement for p300/CBP and PML and is sufficient to stimulate new telomere DNA synthesis. Thus, we demonstrate the application and exploitation of BLOCK-ID to interrogate chromatin responses to replicative stress, while also identifying the epigenetic reader protein, TRIM24, in directing telomere extension by the ALT mechanism.

Results

BLOCK-ID identifies bromodomain proteins at obstructed replication forks

We developed BLOCK-ID (Biotinylation of LacO array repliCation stress protein networK Identification) to detect proteins that accumulate at obstructed replication forks without needing DNA synthesis-dependent nucleoside labeling^4,5,7. In the human osteosarcoma U2OS cell line harboring a 256 repeat lac operator (LacO) array (U2OS 256x), replication fork progression can be blocked by binding of the lac repressor (LacI) to the LacO array^21,22. We fused the promiscuous biotin ligase BirA*²³ to LacI (BirA*-LacI) to enable biotin labeling of proteins associated specifically with blocked replication forks at the LacO array (Figure. 1A). To avoid persistent replication fork blocking, BirA*-LacI expression was controlled by doxycycline. Induction of BirA*-LacI and supplementation of media with biotin led to protein biotinylation, both in bulk measurements by Western blotting and within foci where BirA*-LacI localized to the LacO array (BLOCK-ID foci) (Figure 1B). By immunofluorescence (IF), the accumulation of phospho-ATR kinase (pATR - S428), BRCA1, RPA2, and γH2AX indicated replication stress responses at BLOCK-ID foci (Figure 1C; see methods). Further, DNA damage at the BLOCK-ID locus was suppressed in non-cycling cells but highly enriched in EDU-positive cells, confirming replication-dependent generation of DNA damage (Figure S1A–B).

Figure 1: Stressed replication fork profiling by BLOCK-ID identifies TRIM24.

(A) Schematic of BLOCK-ID. The yellow circle indicates the biotinylation range of BirA* (10 nm). Biotinylated proteins are isolated by biotin affinity pulldown (AP) and analyzed by liquid chromatography mass spectrometry (LC/MS). (B) Western blot of proteins biotinylated by BLOCK-ID. Cells were treated with doxycycline (Dox) and/or biotin (50 μM) for 14 h. Right: representative immunofluorescence (IF) images by confocal microscopy showing the localization of biotinylated proteins at Myc-BirA*-LacI, i.e. BLOCK-ID foci. (C) Replication stress markers localized at BLOCK-ID foci. Statistical significance was determined with unpaired two-tailed T-test, n=3, >100. (D) Gene Ontology (GO) analysis of BLOCK-ID identified proteins. (E) Plot depicting the relative abundance of bromodomain (BRD) proteins identified by BLOCK-ID. (F) Endogenous TRIM24 localized at BLOCK-ID foci. Statistical significance was determined with unpaired two-tailed T-test, n=3, >100. (G) Localization and quantification of endogenous TRIM24 at telomeres. Telomeres were detected using TRF2 antibodies. Cells were treated with 9 μM of RO3306 (G2 arrest) or DMSO for 16 h. Statistical significance was determined with unpaired two-tailed T-test, n=3, >100. (H) Schematic of telomere-specific biotinylation by TurboID-TRF1 in U2OS cells. (I) Western blot of lysates following streptavidin pull-down from U2OS cells expressing TurboID-TRF1 detects TRIM24. (J) Endogenous TRIM24 localization at TRF1-FokI WT or D450A labeled telomeres in U2OS cells. Statistical significance was determined with unpaired two-tailed T-test, n=3, >100. Error bars represent mean ± s.e.m., with P values included in the graphs unless otherwise indicated. NS = Non-significant. Dots on the graphs indicates number of biologically independent experiments and indicated number represents total number of cells or treatments analyzed per each experiment. Representative images are provided and statistical analyses were performed as indicated. Scale bars = 10 μm.

Next, we performed a BLOCK-ID mass spectrometry-based screen. Here, experiments in non-replicating growth-arrested cells and cells expressing BirA* alone were included to control for non-specific biotinylation locally at LacO/I repeats and throughout the cell. After filtering out proteins identified in these controls, we identified a total of 591 proteins (Table S1 and methods), including DNA replication (Figure S1C) and DNA repair factors (Figure S1D–E) with several previously identified by proteomics of HU-arrested replication forks^4,7 (Figure S1F–H). This analysis shows the utility of BLOCK-ID as a method for identifying proteins associated with blocked replication forks. Gene Ontology (GO) analysis also identified DNA and chromatin proteins (Figure 1D). Of those, proteins containing the bromodomain (BRD) were highly enriched, constituting 55% of all ubiquitously expressed human BRD-containing proteins (22 out of 40; Figure 1E). BRD proteins act as “epigenetic readers” of acetylated lysines within histone and non-histone proteins²⁴. Though implicated in DNA repair and replication pathways^25–31, the involvement of BRD proteins in replication stress responses remains poorly understood³². We assessed the recruitment of BRD proteins to BLOCK-ID foci by IF, where 12 out of the 22 tested BRD proteins displayed robust recruitment (Figure S2A; quantified in S2B). Furthermore, depleting 10 of 22 BRD proteins led to >30% cell death in conditions of replication stress invoked by HU treatment (Figure S2C, the efficiency of siRNA depletion is provided in S2D). This analysis further identified BAZ1B, BRD9, TRIM24 and ZMYND8 as putative replication stress responders, showcasing the use of BLOCK-ID in identifying chromatin modifiers involved in replication stress responses.

Mobilization and function of TRIM24 at blocked replication forks

TRIM24 is a member of a large TRIM protein family characterized by a conserved N-terminus sequentially comprised of a RING finger domain, B-box type zinc finger domains, and a coiled-coil (CC) domain³³. TRIM24 also features a tandem plant homeodomain (PHD) and BRD reader domain in the C-terminus. TRIM24 is a transcription coregulator³⁴, a ubiquitin E3 ligase for p53³⁵, and a DNA double-strand break repair-associated factor^26,29. Furthermore, elevated TRIM24 expression is correlated with cancer progression^34,36,37. Using a validated antibody, endogenous TRIM24 localization at BLOCK-ID foci was confirmed (Figure 1F; antibody specificity validated in Figure S3A–B). We also performed SIRF (in situ protein interactions at nascent and stalled replication forks)^28,38 with BLM as a positive control that detected TRIM24 at hydroxyurea (HU) stalled replication forks (Figure S3C). Altogether, these data are consistent with TRIM24 being recruited to stalled replication forks, where it participates in replication stress responses, although we note that formation of DSBs during these treatments may also influence TRIM24 behavior and function.

TRIM24 associates with telomeres in ALT cancer cells

To understand how TRIM24 participates in the DNA replication stress responses, we identified protein-interacting partners of TRIM24 using a proximity ligation approach (i.e., BioID). Employment of BioID using BirA* fused to TRIM24 (BirA*-TRIM24) followed by proteomic analysis identified 188 putative TRIM24 interacting proteins (Table S2), including the established interactors TRIM28³⁹, TRIM33³⁹, and MCM4⁴⁰. GO pathway enrichment analysis of the TRIM24 interactome revealed DNA synthesis, chromatin modification and DNA repair pathways (Fig. S3D–E). However, a subsequent analysis for the subcellular localization of proteins within the TRIM24 interactome revealed that 36% of unique peptides were associated with telomeres, PML bodies, centromeres, Cajal bodies and the spliceosome, with peptides from telomere and PML-associated proteins being the most abundant (Fig. S3F; representation of representative nuclear body TRIM24 interactors).

Based on these observations, we speculated that TRIM24 may associate with telomeres in cells that utilize Alternative Lengthening of Telomeres (ALT) mechanisms. ALT is a specialized homology-directed repair (HDR) pathway that elongates telomeres in cancer cells when telomerase is absent. Studies have established that ALT is stimulated by DNA breaks and replicative stress during G2-M cell cycle phases. Telomere synthesis within ALT cells also occurs in ALT associated PML bodies (referred to as APBs), components of which we identified as putative TRIM24 interactors. Finally, our TRIM24 BioID experiments were performed in the ALT-positive U2OS cell line. Taken together, these findings suggested a potential involvement of TRIM24 in ALT. Indeed, immunofluorescence (IF) analysis of TRIM24 revealed localization to telomeres in cells arrested in G2 with CDK1 inhibitors in ALT+ U2OS cells (Figure 1G). TRIM24 was detected by Western blotting post proximity-dependent biotinylation using telomere-specific TRF1-TurboID (Fig 1H–I). TRIM24 localization to telomeres in U2OS cells was enhanced following the generation of telomere-specific DNA breaks with TRF1-FokI (Figure 1J) and by pyridostatin (PDS) (Figure S4A), a G4-quadruplex stabilizing ligand known to perturb telomere replication⁴¹. Neither endogenous TRIM24 nor overexpressed GFP-TRIM24 were observed at telomeres in asynchronous, G2-arrested, telomerase-positive HeLa cells with long telomeres (HeLa LT) or TRF1-FokI expressing HeLa LT, MG-63 or SJSA1 cells that each express telomerase (Figure S4A–D). The selective and stress-induced localization of TRIM24 with telomeres led us to hypothesize that it could act to mediate the ALT mechanism.

TRIM24 is required for productive ALT

By combining IF for PML with telomeres, we observed that TRIM24 depletion strongly and consistently reduced the frequency of ALT-associated PML bodies in U2OS, SaOS2 and Cal72 ALT cancer cell lines (Figure 2A and Figure S4E–G). We next used a native FISH-based ssTelo assay that detects C-rich telomeric ECTRs²⁰. Here, TRIM24 depletion markedly reduced the number of ssTelo signals in U2OS cells, SaOS2 and Cal72 cells (Figure 2B and Figure S4H–I), being comparable to BLM removal^18,20,42,43, an essential ALT mediator that suppresses aberant telomere recombination. By performing Chromosome-Orientation FISH (CO-FISH) assay on metaphase chromosomes, we found that, unlike BLM deficiency, which elicits high rates of telomere recombination and telomere-sister chromatid exchanges (t-SCEs) (Figure 2C, indicated with carets), TRIM24 depletion decreased t-SCE frequency (Figure 2C). However, the increased t-SCE phenotype of BLM-depleted cells was suppressed after co-depletion of TRIM24 (Figure 2C). Finally, depletion of TRIM24 resulted in reduced cell proliferation in U2OS cells (Figure S4J–L). Taken together, these data implicate TRIM24 in several aspects of telomere functions in ALT including APB formation, telomere recombination, and cell survival.

Figure 2: TRIM24 promotes ALT telomere maintenance.

(A) APBs in U2OS cells treated with CTRL or TRIM24 siRNAs. TRF2 indicates telomeres. Statistical significance was determined using unpaired two-tailed T-test, n=3, >2500. (B) ssTelo foci analyzed in U2OS cells following TRIM24 or BLM depletion using siRNAs. Western blot shows siRNA-depletion of the indicated proteins. Lamin C is the loading control. Statistical significance was determined using one-way ANOVA, n=3, >450. (C) t-SCE analysis by CO-FISH following TRIM24, and/or BLM depletion in U2OS cells. Carets indicate t-SCEs. Statistical significance was determined using one-way ANOVA, n=3, >30 metaphase spreads. Western blot shows depletion of the indicated proteins using siRNAs. (D) Nascent telomere synthesis in U2OS cells following TRIM24 depletion. EdU foci localized to TRF1-FokI labeled telomeres indicate nascent telomere DNAs. Statistical significance was determined using unpaired two-tailed T-test, n=3, >150. (E) Nascent telomere synthesis in U2OS cells following TRIM24 depletion. TRF2 indicates telomeres. EdU foci localized to TRF2-labeled telomeres indicates nascent telomere DNA. Statistical significance was determined using two-tailed T-test, n=3, >290. (F) Telomere length analysis by PFGE. TRIM24-depleted U2OS and HeLa LT cells were generated by stable expression of two independent shRNAs (C1 and C2) targeting the coding region of TRIM24. SCR is non-targeting shRNA. Ethidium bromide (EtBr) is the loading control. Red dots indicate the mean telomere length. (G) Analysis of fragile telomeres and signal free ends in siCTRL and siTRIM24 U2OS cells using telomere FISH on metaphase spreads. Cells from (f) were analyzed. Carets indicates fragile telomeres and asterisks indicate signal free ends. Statistical significance was determined using unpaired two-tailed T-test, n=2, >30 metaphase spreads. Error bars represent mean ± s.e.m., with P values included in the graphs unless otherwise indicated. NS = Non-significant. Dots on the graphs indicates number of biologically independent experiments and indicated number represents total number of cells or treatments analyzed per each experiment. Representative images are provided and statistical analyses were performed as indicated. Scale bars = 10 μm.

To determine the impact of TRIM24 deficiency on nascent telomere DNA synthesis, we monitored Ethynyl-2’-deoxyuridine (EdU) labeling at telomere DNA breaks induced by TRF1-FokI. This method induces “ALT-like” telomere processes including new telomere DNA synthesis¹². Without TRIM24, we observed ~50% fewer telomeres with focal accumulation of EdU (Figure 2D). We observed a similar defect in telomere synthesis in G2-arrested U2OS cells depleted for TRIM24 (Figure 2E). In stably depleted TRIM24 U2OS and HeLa LT cell lines (Figure S5A), we measured telomere length by pulsed-field gel electrophoresis of telomere restriction DNA fragments followed by Southern blotting with telomere probes. This analysis revealed substantial telomere shortening in TRIM24-deficient U2OS cells but not in TRIM24-deficient HeLa LT cells (Figure 2F). Similarly, metaphase chromosome FISH analysis revealed a greater frequency of telomere loss in TRIM24 deficient cells (Figure 2G, indicated with asterisks). Moreover, telomere fragility, a phenotype linked with defects in telomere replication, was markedly increased in TRIM24-deficient U2OS cells (Figure 2G, indicated with carets). Accordingly, TRIM24-deficient U2OS harbored elevated RPA2 phosphorylation (phospho-RPA-S33) (Figure S5A) despite no appreciable alteration in cell cycle progression (Figure S5B). Importantly, these substantial alterations in telomere length and ALT-associated telomere DNA synthesis were observed in an independent LM216J ALT cell line, but not its isogenic LM216T partner that expresses telomerase (Figure S5C–F). TRIM24-deficient LM216J cells also exhibited elevated pRPA33, impaired APB formation and defective nascent telomere DNA synthesis (Figure S5C,G–H). This comprehensive analysis of ALT markers and telomere length demonstrates that TRIM24 promotes ALT-dependent telomere extension.

CBP and p300 recruit TRIM24 to telomeres to facilitate ALT

Using a series of domain deletion derivatives of TRIM24, the localization of TRIM24 to telomeres was determined to be dependent on the PHD and BRD domains (Figure 3A). We also tested TRIM24 WT and mutants upon overexpression, observing that WT TRIM24 overexpression resulted in increased APB and ssTelo formation (Figure S5I–K). Deletion of the PHD domain suppressed ssTelo formation, highlighting a role for this reader domain in ALT under conditions where TRIM24 is overexpressed (Figure S5I–K). The PHD and BRD of TRIM24 are chromatin “reader” domains, which have been shown to bind a chromatin signature of unmethylated H3K4 and acetylated H3K14 or H3K23 to direct gene regulation in breast cancer cells^37,44. We reasoned that these histone modifications may mediate the interaction between chromatin and TRIM24 at telomeres in ALT cancer cells. We first examined whether depleting any of 12 major histone acetyltransferases (HATs) altered TRIM24 recruitment to TRF1-FokI telomeric breaks. Here, we found that depleting CBP and/or p300 reduced the localization of TRIM24 to TRF1-FokI foci by nearly 50% (Figure 3B; Screen shown in Figure S6A–C). TRIM24 localization was similarly ablated at TRF1-FokI foci after treating cells with the CBP and p300 HAT activity inhibitor, A-485⁴⁵ (Figure S7A). Depletion of another H3K23 HAT, MORF, or its associated protein BRPF1, did not affect TRIM24 localization to telomeres⁴⁶ (Figure S6A–F). Importantly, depleting CBP or p300 did not alter the mRNA or protein level of TRIM24 (Figure S7B–C), suggesting a direct role for the HAT activity of CBP and p300 in directing TRIM24 to telomeres in ALT cancer cells. Indeed, CBP and p300 localized to TRF1-FokI induced telomere DSBs in U2OS cells (Figure 3C) but not in nonALT, telomerase-positive HeLa LT cells (Figure S7D).

Figure 3: CBP and p300 direct TRIM24 to telomeres and promote ALT.

(A) Localization of GFP tagged TRIM24 WT and deletion variants to TRF1-FokI labeled telomeres in U2OS cells. Scheme of WT TRIM24 and deletion variants is shown (top right). Statistical analyses were performed with one-way ANOVA, n=3, >100. (B) Endogenous TRIM24 localized at TRF1-FokI labeled telomeres in U2OS cells following CBP and/or p300 depletion. Western blot shows depletion of the indicated proteins using siRNAs. Lamin C is the loading control. Statistical analyses were performed with one-way ANOVA, n=3, >100. (C) Endogenous CBP and p300 localized at TRF1-FokI WT and D450A foci labeled telomeres in U2OS cells. Statistical significance was determined using unpaired two-tailed T-test, n=3, >105. (D) IF analysis of H3K23ac localized to TRF1-FokI labeled telomeres in U2OS cells following A-485 treatment. Cells were treated with 5 μM of A-485 or DMSO for 16 h. Statistical analyses were performed with two-tailed T-test, n=3, >400. (E) \ ssTelo foci detection in U2OS cells following CBP or p300 depletion. Statistical analyses were performed with one-way ANOVA, n=3, >420. (F) Representative images and quantification of APBs and telomere foci size in U2OS cells treated with A-485 or SGC-CBP30. Cells were treated with 1 μM of A-485 or 1 μM of SGC-CBP30 for 3 days. Telomere size after A-485 or SGC-CBP30 treatment was determined using telomere FISH (TelFISH). Statistical analyses were performed with one-way ANOVA, n=3, >2500 for APBs and >230 for foci size. (G) Nascent telomere synthesis in U2OS cell treated with A-485 or SGC-CBP30. EdU foci localized at TRF1-FokI labeled telomeres indicate nascent telomere DNAs. Cells were treated with 1 μM of A-485 or SGC-CBP30 for 3 days. Statistical analyses were performed with one-way ANOVA, n=3, >150. (H) Fragile telomeres and signal free ends in U2OS cells treated with A-485. Cells were treated with 1 μM of A-485 or DMSO for 3 days. Carets indicate fragile telomeres and asterisks indicates signal free ends. Statistical analyses were performed with two-tailed T-test, n=2, >30 metaphase spreads. (I) Analysis of t-SCEs using CO-FISH following A-485 treatment. U2OS cells were treated with 1 μM of A-485 for 16 h. Statistical significance was determined using unpaired T-test. n=3, >40 metaphase spreads. Error bars represent mean ± s.e.m., with P values included in the graphs unless otherwise indicated. NS = Non-significant. Dots on the graphs indicates number of biologically independent experiments and indicated number represents total number of cells or treatments analyzed per each experiment. Representative images are provided and statistical analyses were performed as indicated. Scale bars = 10 μm.

Using specific antibodies, both H3K23ac and H3K14ac were detected in the biotinylated protein fraction recovered from streptavidin pulldown in U2OS cells expressing TurboID-TRF1. Depleting CBP markedly diminished the level of H3K23ac level in the biotinylated protein fraction, whereas H3K14ac remained unchanged (Figure S7E–F). In agreement with these biochemical data, H3K23ac accumulated at TRF1-FokI sites (Figure S7G), which was diminished after treatment with A-485 (Figure 3D). We then speculated that inhibiting CBP and p300 would replicate the telomere phenotypes observed in TRIM24-deficient cells. Indeed, depleting CBP or p300 decreased ssTelo signals, APB formation and telomere DNA synthesis in U2OS cells (Figure 3F–G). CBP/p300 inhibition also increased both telomere fragility (indicated with carets) and telomere loss (indicated with asterisks) on metaphase chromosomes from U2OS cells while increasing t-SCEs (Figure 3H–I). While CBP/p300 deficiency resulted in defective telomere integrity, similarly to TRIM24 loss, we also analyzed the effects of inhibiting both p300/CBP and TRIM24 in ALT+ cells. For ssTelo and APBs, we found that inhibiting p300/CBP with 1uM for A-485 in TRIM24-depleted cells elicited the same levels of alteration as either single perturbation. p300/CBP inhibition did reduce the proliferation of TRIM24-depleted cells, albeit only at the elevated 5μM dose (Figure S7H–J). Based on these results, TRIM24 and p300/CBP may be epistatic in ALT, although we cannot rule out that additional p300/CBP targets are also involved. Concerning cell proliferation, p300/CBP modifies many proteins and impacts diverse cellular pathways, including transcription. Collectively, these data reveal a chromatin signaling axis involving p300/CBP and TRIM24 that is integral for preserving telomeres in ALT cancer cells.

Constitutive tethering of TRIM24 to telomeres promotes APB formation and enhances ALT

We next queried whether constitutive tethering of TRIM24 at telomeres could, by itself, affect ALT. The DNA-binding domain of the S. pombe protein Teb1 that recognizes human telomere TTAGGG sequences^20,47 was fused to the C-terminus of TRIM24 (TRIM24-TebDB). IF analysis confirmed TRIM24-TebDB telomeres localization and remarkably, TRIM24-TebDB was sufficient to stimulate ALT activities, as indicated by increased ssTelo signals compared to control U2OS cells (Figure 4A–B). CBP/p300 deficiency abolished ssTelo signals consistent with its requirement for ALT. However, telomeric tethering of TRIM24 in either CBP-depleted cells or A-485-treated cells stimulated ssTelo formation (Figure S8A and Figure 4B). These data support the premise that the major function of these HATs in ALT is to promote the localization of TRIM24 to telomeres.

Figure 4: TRIM24 promotes APB formation at ALT telomeres.

(A) Localization of Flag-TebDB and Flag-TRIM24-TebDB localized to telomeres in U2OS cells. TRF2 indicates telomeres. Illustration of Teb1 and Flag-TRIM24-TebDB is shown (top). Teb1 DNA binding domain (TebDB) was fused to the c-terminus of TRIM24. Statistical analyses were performed with two-tailed T-test, n=3, >75. (B) Representative images and quantification of ssTelo foci analyzed in U2OS cells either treated or not-treated with 5 μM of A-485 for 72 h with or without expressing GFP-TRIM24-TebDB. Statistical significance was determined using one-way ANOVA, n=3, >100. (C) Recruitment of endogenous PML by TebDB or TRIM24-TebDB in U2OS cells. Statistical analyses were performed with two-tailed T-test, n=3, >100. (D) HA-PML IV, GFP-MRE11, Myc-NBS1, and GFP-BLM recruitment by Flag-TRIM24-TebDB in U2OS cells. Statistical analyses were performed with two-tailed T-test, n=3, >100. (E) Endogenous TRIM24 localized at TRF1-FokI labeled telomeres in U2OS cells following PML depletion using siRNAs (siPML). Statistical analyses were performed with two-tailed T-test, n=3, >125. (F) Representative images of clustered telomeres in either mock-treated or HA-PML IV-TebDB (PML-TebDB) expressing U2OS cells following TRIM24 depletion using siTRIM24. GFP-TRF1 marks telomeres. Schematic of HA-PML IV fused to TebDB (PML-TebDB) is shown (top right). Quantification of telomere size with one-way ANOVA, n=3, >60. (G) Domain structure of WT TRIM24 fused to TebDB and deletion variants. (H) Representative images of HA-PML IV recruited by TRIM24-TebDB variants depicted in (G). (I) Quantification of (H). Statistical analyses were performed with one-way ANOVA, n=3, >110. (J) Telomere clustering induced by TRIM24-TebDB variants depicted in (G). GFP-TRF1 labels telomeres. Size of clustered telomeres indicated below. Statistical significance was determined by one-way ANOVA, n=3, >69. Error bars represent mean ± s.e.m., with P values included in the graphs unless otherwise indicated. NS = Non-significant. Dots on the graphs indicates number of biologically independent experiments and indicated number represents total number of cells or treatments analyzed per each experiment. Representative images are provided and statistical analyses were performed as indicated. Scale bars = 10 μm.

We reasoned that this system would provide a powerful experimental tool to dissect the sequence of molecular events dictated by p300-CBP-TRIM24 in ALT. Given the interaction between TRIM24 and PML, and the role of PML in forming APBs^20,43, we tested the ability of TRIM24-TebDB to recruit PML to ALT telomeres. While TebDB alone displayed telomere localization, including at a few PML foci, tethering of TRIM24-TebDB resulted in a near-complete recruitment of PML to telomeres (Figure 4C). In addition to PML, several other APB constituents, including NBS1, MRE11, and BLM were also highly recruited to TRIM24-TebDB sites (Figure 4D). Immunoprecipitation of TRIM24 also captured an interaction with BLM, which is consistent with our TRIM24-BioID data and further supports TRIM24 interactions with known APB components (Figure S8C). Previous studies identified that the fourth splicing isoform of PML (PML IV) is essential for ALT^20,43. We found that TRIM24 and HA-PML-IV interacted by co-immunoprecipitation (Figure S8B). These data indicate that TRIM24 interacts with several factors involved in APBs and that TRIM24 localization at telomeres stimulates the mechanism that establishes bona fide APBs.

To further understand this relationship with APB formation, we first depleted PML, which did not affect telomere-localization of endogenous TRIM24 (Figure 4E and Figure S8D). Consistent with previous observations⁴³, tethering PML-IV to telomeres significantly increased the size of telomeric foci, marked by GFP-TRF1, in U2OS cells (Figure 4F). However, tethering PML-IV to telomeres failed to induce enlarged telomere foci in TRIM24-depleted U2OS cells (Figure 4F), suggesting that TRIM24 may indeed be required for telomere clustering within APBs. To decipher how TRIM24 is involved in telomere clustering, we expressed several TRIM24-TebDB deletion mutants (Figure 4G). Only the deletion of the coiled-coil domain (CC) of TRIM24 abolished PML-IV telomere recruitment and telomere clustering as measured by TRF1-GFP size, despite comparable levels of transgene expression (Figure 4G–J; Figure S8E). Thus, the CC domain of TRIM24 is critical for APB formation and telomere clustering.

SUMOylation of TRIM24 promotes the formation of functional APBs

SUMOylation is a major driver of APB formation^20,42,48. Western blot analysis of GFP-TRIM24 readily detects several slowly migrating protein species (Figure 5A), previously characterized as SUMOylated TRIM24⁴⁴. Bands representing SUMOylated TRIM24 were not detected when TRIM24 lacked either the PHD or BRD (TRIM24ΔPHD or TRIM24ΔBRD, respectively), as well as in SUMO-Defective TRIM24 (TRIM24-SD); with lysine to arginine mutations at the major SUMOylation sites K723R and K741R⁴⁴ (Figure 5A). Interestingly, the level of chromatin-bound SUMO-TRIM24 was enhanced by TRF1-FokI and treatment with the HDAC inhibitor Trichostatin A (TSA) in U2OS cells (Figure S8F). Thus, these results support the involvement of both SUMOylation and histone acetylation in regulating TRIM24 telomere localization and subsequent function in ALT cancer cells.

Figure 5: Analysis of TRIM24 in PML-dependent and -independent ALT.

(A) Schematic of GFP-tagged WT TRIM24 and mutants (top). SUMOylation deficient (SD) TRIM24 has K to R mutations in amino acids 723 and 741. Western blot of lysates expressing WT TRIM24 and derivatives using anti-GFP and Tubulin as loading control is shown (bottom). Slower migrating SUMOylated forms of TRIM24 are indicated. (B) APBs analyzed in U2OS cells treated with siCTRL or siTRIM24 3’ UTR followed by complementation with the indicated TRIM24 construct. TRF2 marks telomeres. Cells were treated with 9 μM of RO3306 for 16 h prior to analysis to enrich for G2 cells. Statistical analyses were performed with one-way ANOVA, n=3, >100. (C) Nascent telomere DNA synthesis determined in U2OS cells either treated siCTRL or siTRIM24 3’ UTR followed by complementation with the indicated TRIM24 constructs. EdU foci localized at TRF1-FokI labeled telomeres indicate telomeres. Cells were treated with 9 μM of RO3306 for 16 h. Statistical analyses were performed with one-way ANOVA, n=4, >180. (D) ssTelo foci analyzed in mock-treated or Flag-TRIM24-TebDB expressing U2OS cells following PML-depletion using siRNAs. Statistical analyses were performed with one-way ANOVA, n=3, >100. (E) Nascent telomere DNA synthesis analyzed in parental or PML KO U2OS cells expressing Flag-TRF1 or Flag-TRIM24-TRF1 fusion construct. Cells were treated with 9 μM of RO3306 for 16 h. Statistical analyses were performed with two-tailed T-test, n=3, >125. Error bars represent mean ± s.e.m., with P values included in the graphs unless otherwise indicated. NS = Non-significant. Dots on the graphs indicates number of biologically independent experiments and indicated number represents total number of cells or treatments analyzed per each experiment. Representative images are provided and statistical analyses were performed as indicated. Scale bars = 10 μm.

To dissect the mechanistic connection between SUMO and TRIM24 during ALT, we found that treating U2OS cells expressing TRIM24-TebDB with TAK-981, a potent inhibitor of SUMO-activating enzyme subunit 2 (SAE2)⁴⁹, abrogated the recruitment of PML (either endogenous or ectopically expressed PMLIV) to TRIM24-TebDB sites (Figure S8G–I). To directly investigate the role of TRIM24 SUMOylation in APB formation, we examined the abundance of telomere-localized PML in TRIM24-deficient cells complemented with TRIM24 mutants (Figure 5A). As before, APB formation was reduced upon TRIM24 depletion (Figure 5B) in G2-arrested U2OS cells. Expressing WT TRIM24, but neither TRIM24ΔCC or TRIM24-SD, restored the reduced abundance of APBs or telomere DNA synthesis defects observed in TRIM24-depleted cells (Figure 5B–C and Figure S8J). Taken together, these findings demonstrate that TRIM24 SUMOylation, as well as its coiled-coil domain, are required to facilitate ALT through an ability to form functional APBs.

Telomere-tethered TRIM24 promotes ALT independent of PML

We wanted to test further the functional consequences of TRIM24-PML interactions in supporting ALT. Consistent with previous reports²⁰, depletion or knockout (KO) of PML attenuated ALT, abolishing ssTelo signals and nascent telomere synthesis in U2OS cells (Figure 5D–E and Figure S9A). Strikingly, using two independent strategies, TebDB or the widely used TRF1 protein, tethering TRIM24 at telomeres restored telomere synthesis and ssTelo signals in PML-deficient U2OS cells (Figure 5D–E and Figure S9A). TRIM24-TebDB was also sufficient to recruit BLM to these sites in both control and PML-deficient cells (Figure S9B). This was surprising since PML is considered indispensable for recruiting BLM to telomeres in ALT cancer cells²⁰. In WT U2OS cells, TRIM24 depletion reduced BLM localization at telomeres in G2-arrested cells (Figure S9C). These results are congruent with TRIM24 promoting BLM localization to telomeres and previous findings that telomere tethered BLM-TOPIIa-RMI1 (BTR) complex induced ALT activity in PML KO cells²⁰. In line with BLM playing an essential role in ALT, BLM-deficiency abolished ssTelo formation by TRIM24-TebDB in PML-depleted cells (Figure S9D–E). Thus, tethering TRIM24 to telomeres does not bypass the essential requirement for BLM in ALT, unlike PML. Similarly, BLM recruitment to TRIM24-TebDB was diminished by TAK-981 treatment (Figure S9B), indicating that BLM recruitment by TRIM24-TebDB is SUMOylation-dependent. Taken together, our data suggests that artificially tethering TRIM24 to telomeres is alone sufficient to induce ALT-like activities in a PML-independent but BLM- and SUMOylation-dependent manner.

TRIM24 collaborates with MMS21 and PIAS1 to promote ALT

Lastly, we sought to identify the SUMO E3 ligase(s) involved in this pathway, focusing on PIAS1, PIAS4 and MMS21 due to their established roles in the DDR⁵⁰ and ALT⁵¹. First, we observed that the SUMOylation of chromatin-bound TRIM24 was reduced upon the depletion of either PIAS1 or MMS21, but not PIAS4 (Figure 6B; siRNA-depletion of proteins shown in Figure S9F). Depletion of PIAS1 or MMS21 noticeably decreased ssTelo signals (Figure 6C). We speculated that SUMOylation of TRIM24 might regulate its recruitment and/or retention at telomeres. In contrast to WT-TRIM24, SUMO-defective TRIM24-SD localization to telomeres was impaired (Figure 6D). Furthermore, depleting MMS21, but not PIAS1, significantly reduced the telomere localization of TRIM24 (Figure 6D). However, upon TRIM24 tethering to telomeres, PIAS1-depletion reduced TRIM24 SUMOylation while MMS21-deficiency had no effect (Figure 6F). These data indicate that once TRIM24 is localized to telomeres, TRIM24 SUMOylation is dependent on PIAS1. In line with the involvement of PIAS1 in mediating TRIM24 SUMOylation at telomeres, expression of TRIM24-TebDB robustly recruited PIAS1 but not MMS21 to these sites, which was dependent on the CC domain of TRIM24 (Figure S9G–H). Functionally, we observed a substantial reduction in ssTelo signals in TRIM24-TebDB expressing cells that were either PML proficient or deficient (Figure 6G–H), indicating the importance of PIAS1-mediated SUMOylation in promoting ALT activities. These results suggest that MMS21 SUMOylates factors other than TRIM24 to facilitate its recruitment to telomeres, whereupon TRIM24 becomes SUMOylated by PIAS1 through a mechanism dependent on the CC-domain, a pathway that collectively promotes ALT.

Figure 6: SUMOylation regulates TRIM24 to promote ALT.

(A) ssTelo foci analyzed in WT or SD TRIM24-TebDB expressing U2OS cells following siRNA depletion of PML. Statistical analyses were performed with unpaired two-tailed T-test, n=3, > 95. (B) Western blot of WCE (whole cell extract) and Chr (chromatin fraction) from U2OS cells treated with the indicated siRNAs was performed to detect SUMOylation of endogenous TRIM24. Tubulin and H3 serve as loading controls for cytoplasmic fraction and chromatin fraction, respectively. (C) ssTelo foci analyzed in U2OS cells following siRNA treatments targeting CTRL, PIAS1 or MMS21. Statistical analyses were performed with one-way ANOVA, n=3, >200. (D) GFP-tagged TRIM24 WT and SD localization to TRF1-FokI labeled telomeres in TRIM24-depleted U2OS cells. Statistical analyses were performed with unpaired two-tailed T-test, n=2, >60. (E) Endogenous TRIM24 localized to TRF1-FokI labeled telomeres in U2OS cells following PIAS1 or MMS21 depletion. Statistical analyses were performed with one-way ANOVA, n=3, >110. (F) Western blot of lysates from WCE and Flag purified immunocomplexes with anti-Flag and SUMO2/3 to detect SUMOylation of Flag-TRIM24-TebDB in U2OS cells following PIAS1- or MMS21-depletion by siRNAs. (G) ssTelo foci analyzed in TRIM24-TebDB expressing PML-depleted U2OS cells following PIAS1-depletion. Statistical analyses were performed with unpaired two-tailed T-test, n=3 >75. (H) ssTelo foci in TRIM24-TebDB expressing parental U2OS cells following PIAS1-depletion. Statistical analyses were performed with unpaired two-tailed T-test, n=2, > 65. Error bars represent mean ± s.e.m., with P values included in the graphs unless otherwise indicated. NS = Non-significant. Dots on the graphs indicates number of biologically independent experiments and indicated number represents total number of cells or treatments analyzed per each experiment. Representative images are provided and statistical analyses were performed as indicated. Scale bars = 10 μm.

This comprehensive analysis defines a sequence of molecular events involving TRIM24 that preserves telomeres in ALT cancer cells. First, histone acetylation by CBP/p300 and SUMOylation activity by MMS21 promote the localization of TRIM24 to telomeres where it relies on PIAS1-mediated TRIM24 SUMOylation and BLM to connect telomeres to APBs for telomere DNA synthesis (Figure 7). In PML-deficient cells, we uncovered that merely tethering TRIM24 at telomeres is sufficient to stimulate the assembly and activity of APB-like bodies (Figure 7). Crucially, disrupting this pathway abrogates ALT activity implicating this CBP/p300-TRIM24-SUMO axis in directing critical processes required for telomere length maintenance in ALT cancer cells.

Figure 7: Model for TRIM24 Regulation and Function in ALT.

The bromodomain protein TRIM24 promotes both PML-dependent APBs and ALT-associated PML-independent bodies to facilitate telomere maintenance through ALT. CBP and p300 acetylate H3K23ac to direct TRIM24 to telomeres in a SUMOylation dependent pathway involving MMS21. Once localized to ALT telomeres, TRIM24 undergoes PIAS1-dependent SUMOylation leading to the formation of both APBs and ALT-associated PML-independent bodies that promote ALT.

Discussion

In this study, we developed the BLOCK-ID technique, which combines unbiased mass spectrometry and single-cell fluorescence microscopy imaging of protein dynamics of stressed replication forks. BLOCK-ID harnesses an engineered genomic locus containing the Lac Operator (LacO) repeat array that when bound by the Lac Repressor (LacI) forms a protein complex barrier that impedes replication fork progression. While several powerful approaches have been developed to detect replication stress-associated proteins, including iPOND⁵ and Nascent Chromatin Capture⁷, BLOCK-ID does not require nucleotide labeling, instead enabling direct detection of proteins at stressed replication forks by proximity-dependent biotinylation (Figure 1A). This powerful feature of BLOCK-ID preserves the capacity to capture proteins that are transiently and/or weakly associated with stressed replication forks. Comparisons of proteins enriched between BLOCK-ID, iPOND and NCC revealed nice concordance between BLOCK-ID and these established methodologies. The utility of BLOCK-ID to capture chromatin modifiers was validated by identifying nearly half of the 42 human Bromodomain proteins (Figure 1E), highlighting the importance of this chromatin reader protein family in DNA replication responses³².

Here, we focused on TRIM24, which we characterize as a replication stress response factor with an unexpected but essential role in telomere extension by the ALT mechanism. We determined that TRIM24 localizes to telomeres in ALT cancer cells where it contributes to replicative stress mitigation and telomere structural integrity (Figure 1G, and Figure 2F). Furthermore, disruption of TRIM24 robustly interferes with telomere DNA synthesis (Figure 2D–E), leading to telomere destabilization and rapid shortening (Figure 2F). Mechanistically, we identified that p300 and CBP-dependent H3K23 acetylation mediates TRIM24 recruitment to damaged telomeres (Figure 3A–D). The PHD-BRD of TRIM24 forms contacts with each other in the crystal structure and mutations in either the PHD or BRD disrupt binding to H3K23ac³⁷. Accordingly, TRIM24 mutants lacking these reader modules did not associate with telomeres (Figure 3A). We found that once TRIM24 is recruited to telomeres, CBP/p300 is dispensable for downstream ALT activities (Figure 4B and Figure S8A) suggesting that one of the main functions of p300/CBP in ALT is to modulate telomeric chromatin to allow TRIM24 to engage with telomeres. However, p300/CBP likely has other targets at telomeres that act upstream of TRIM24 recruitment to telomeres in ALT. For example, TRF2 is acetylated by p300⁵², which may also contribute to TRIM24 accumulation at telomeres and ALT. In addition, inhibition of p300/CBP does not entirely abolish TRIM24 recruitment to telomeres, suggesting that there may be other factors that regulate this pathway.

In addition to acetylated H3, TRIM24 also binds unmodified H3K4 and methylated H3K9, a repressive histone mark associated with heterochromatin, including at stalled replication forks where it protects against fork degradation⁸. The dual PTM reader function of TRIM24 with acetylation and methylation marks could enable TRIM24 to stably associate at telomeres whereby it is initially recruited via H3K23ac and then retained through H3K9me3 binding. Notably, SETDB1 mediated H3K9me3 at telomeres⁵³, which like TRIM24, is implicated in establishing and maintaining productive ALT. In mouse embryonic stem cells, TRIM24 has been shown to simultaneously bind to both p53 and unmethylated H3K4 to promote the opening of condensed chromatin for transcriptional activation⁵⁴. Thus, open and closed chromatin confirmations governed by multivalent histone mark engagement may regulate the function of TRIM24 in diverse biological processes including transcription, and as demonstrated here, ALT telomere maintenance. Future work could dissect TRIM24s chromatin interactions at telomeres specifically, including possible links to SETDB1-dependent H3K9me3, to elucidate further the functional contributions of TRIM24 with ALT.

In addition to the CBP/p300 network that regulates TRIM24 recruitment to telomeres, the relationship between TRIM24 and ALT-associated PML bodies was particularly striking. We initially found that disrupting TRIM24 impaired APB formation. By artificially tethering TRIM24, through fusion with TebDB or TRF1, we observed the striking de novo enrichment of PML protein to all telomeres (Figure 4C–D). Through deletion studies, we uncovered this activity requires the SUMOylation of TRIM24 by PIAS1 (Figure 6B, and 6F), which facilitates interactions with other key mediators of ALT such as the BLM helicase (Figure S9B). While telomere-tethered TRIM24 can recruit BLM to telomeres, we found that TRIM24 depletion resulted in the loss of BLM localization to ALT telomeres (Figure S9C). This positions TRIM24 upstream of BLM in the mechanism of APB formation. It is tempting to speculate that SUMOylated TRIM24 promotes the telomere-localization of BLM and PML, both of which bind to SUMO through SIM motifs^42,55. Therefore, TRIM24 recruitment at telomeres might seed and bridge the formation of APBs It is worth noting that TRIM24 was previously implicated in aberrant PML dynamics. The fusion oncoprotein PML-RARα, which forms dysregulated interactions with TRIM24⁵⁶, impairs the proper formation of PML nuclear bodies, leading to the dysregulation of BLM and subsequent defects in homologous recombination⁵⁷. This implicates, along with our data provided here, the role of TRIM24 as a regulatory factor for the functional formation of PML nuclear bodies, including at telomeres.

We also found that telomere tethering of TRIM24 was sufficient to induce de novo synthesis of telomeric DNA, in cells devoid of PML (i.e. PML KO cells) (Figure 5D–E). Our findings indicate that TRIM24 cooperates with BLM and SUMO to stimulate PML-independent ALT activity. How such PML-independent telomere DNA synthesis is initiated is not known. BLM was shown to initiate ALT-telomere DNA synthesis at misprocessed Okazaki fragments⁵⁸. Tethering of SUMO3 using a chemical dimerizer can also promote ALT features by recruiting BLM, RAD51AP1, and RAD52⁵⁹. Among these factors, BLM recruitment to telomeres by SUMO3 induces ALT in PML KO cells in a SUMOylation-dependent manner⁵⁹. This is consistent with our findings that mutation of two key SUMOylation sites on TRIM24 reduced its ability to promote ALT, highlighting the involvement of SUMOylation in ALT. Determining how TRIM24 can bypass the requirement for PML should be explored further. Identifying SUMO-dependent interactors of TRIM24 specifically in PML-deficient ALT cells at telomers may provide key insights into the molecular mechanisms regulating APIBs formation that we report here. Regardless, the fact that nascent telomere DNA synthesis can be stimulated in the absence of PML by TRIM24 is an important observation, raising further questions about the functional relationship between PML and TRIM24.

In summary, we have identified a p300/CBP/TRIM24 chromatin signaling pathway that promotes ALT telomere maintenance via BLM and PIAS1-mediated SUMOylation (Figure 7). Our mechanistic findings highlight a previously uncharacterized role for acetylation signaling and TRIM24 in ALT. The finding that TRIM24 recruits PML to telomeric chromatin provides unexpected insights and a mechanistic basis for the known telomere localization and DNA synthesis of telomeres within APBs in ALT. The strong reduction in ALT activities, including the rapid loss of telomeres, in TRIM24-deficient cells demonstrates its essential function in ALT telomere maintenance. These findings provide support for further studies to evaluate p300/CBP/TRIM24 as therapeutic targets in ALT.

Limitations of the study

BLOCK-ID might not capture the entire landscape of proteins associated with stressed replication forks throughout the genome in different environments. Additionally, the slow kinetics of BirA* might favor identification of proteins associated with collapsed replication forks. However, employing biotin ligases with faster kinetics such as Turbo-ID⁶⁰ in future studies might allow for more temporal investigation of stressed replication forks by BLOCK-ID. Despite these limitations, our study establishes BLOCK-ID as a powerful, broadly applicable approach to interrogate the replication stress response, whose use will likely deliver mechanistic insights of how replication forks are protected and processed from stresses that when impaired drive genome instability and cancer initiation.

While our study identified p300/CBP-mediated H3K23ac in promoting TRIM24 binding to ALT telomeres, these HATs may have other targets in ALT. p300/CBP are known to be promiscuous enzymes with many targets⁶¹, including TRF2 at telomeres⁶¹. p300/CBP contain bromodomains that bind other acetylated lysine residues. Those modifications likely modulate different aspects of the ALT mechanism, including potentially the localization of p300/CBP to ALT telomeres. While we screened for the involvement of HATs in promoting TRIM24 recruitment to ALT telomeres, we cannot rule out that other HATs in addition to p300/CBP may function in ALT. This work provides the impetus to evaluate the functional contribution of histone lysine acetylations, as well as their readers, writers and erasers, in ALT. It is also possible that p300/CBP has additional targets beyond TRIM24 that participate in ALT, which is worth considering given our observation that p300/CBPi and depletion impact ALT activities. Given the recent development of HAT inhibitors, including against p300 and CBP inhibitors⁶², addressing these questions in the future may provide therapeutic strategies to target ALT cancers with epigenetic drugs targeting acetylation pathways including TRIM24 and p300/CBP.

Resource availability

Lead Contact

Request for information and resources should be directed to and will be fulfilled upon completion by the Lead Contact, Kyle M. Miller (kyle.m.miller@emory.edu).

Materials availability

Reagents generated in this study will be made available by the Lead Contact upon appropriate completion of a Material Transfer Agreement, if necessary. Several cell lines were obtained from other labs that are listed in the acknowledgements and where requests should be made.

Data and code availability

• Mass spectrometry data is deposited in ProteomeXchange database with accession numbers provided in the key resources table. Original imaging data files are available on Mendeley Data (DOI: 10.17632/bw5v3pb3vy.1) as of the date of publication.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

KEY RESOURCES TABLE.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-TRIM24 (WB)	Abcam	Cat# ab256491
Anti-TRIM24 (IF)	Santa Cruz	Cat# sc-271266, RRID:AB_10611751
Anti-phosphoATR (S428)	GeneTex	Cat# GTX133852; RRID:AB_2887120
Anti-BRCA1	Santa Cruz	Cat# sc-6954, RRID:AB_626761
Anti-RPA	Abcam	Cat# ab2175, RRID: AB_302873
Anti-pRPA (S33)	Fortis Life Sciences	Cat# A300-246A, RRID:AB_2180847
Anti-γH2AX (S139)	Millipore	Cat# 05-636-AF647, RRID:AB_309864
Anti-γH2AX (S139)	Novus Biologicals	Cat# NB100-384, RRID:AB_350295
Anti-Myc	Abcam	Cat# ab9106, RRID:AB_307014
Anti-Myc (9E10)	OriGene	Cat# TA150121, RRID:AB_2622266
Anti-HA	Abcam	Cat: ab9110, RRID: AB_307019
Anti-CBP	Santa Cruz	Cat# sc-7300, RRID:AB_626817
Anti-p300	Cell Signaling Technology	Cat# 86377S, RRID:AB_2800077
Anti-PML (IF)	Abcam	Cat# ab96051, RRID:AB_10679887
Anti-PML (WB)	Fortis Life Sciences	Cat# A301-167A, RRID:AB_873108
Anti-TRF2	Novus Biologicals	Cat# NBP1-86911, RRID:AB_11036996
Anti-Flag	Sigma-Aldrich	Cat# F7425, RRID:AB_439687
Anti-Flag (M2)	Cell Signaling	Cat# 2368, RRID:AB_2217020
Anti-GFP	Abcam	Cat# ab290, RRID:AB_303395
Anti-β–Tubulin	Abcam	Cat# ab6046, RRID:AB_10807712
Anti-Baz2B	Abnova	Cat#: H00029994-B01P, RRID:AB_1113146
Anti-BRD3	Fortis Life Sciences	Cat#: A302-368A, RRID: AB_1907251
Anti-BRD9	Abcam	Cat#: ab137245
Anti-MLL1	Novus	Cat#: NB600-248, RRID:AB_3193987
Anti-PBRM	Fortis Life Sciences	Cat# A301-591, RRID:AB_1078808
Anti-SP100	Enzo Life Sicences	Cat#: BML-PW0325-0025
Anti-TRIM28	Santa Cruz	Cat# sc-33186, RRID:AB_2209905
Anti-TRIM33	Fortis Life Sciences	Cat# A301-059A, RRID:AB_873150
Anti-Biotin	Fortis Life Sciences	Cat#: A150-109A, RRID:AB_67327
Anti-BRPF1	GeneTex	Cat# GTX102778, RRID:AB_1949759
Anti-BLM (WB)	Fortis Life Sciences	Cat# A300-110A, RRID:AB_2064794
Anti-BLM (IF)	Santa Cruz	Cat# sc-365753, RRID:AB_10851630
Anti-H3	Abcam	Cat# ab1791, RRID:AB_302613
Anti-H3K23ac	Active Motif	Cat# 39131, RRID:AB_2793165
Anti-H3K14ac	Diagenode	Cat# C15210005, RRID:AB_2713906
Anti-PIAS1	Abcam	Cat# ab32219, RRID:AB_777265
Anti-PIAS4	Abcam	Cat# ab58416, RRID:AB_881880
Anti-MMS21	Fortis Life Sciences	Cat# A304-126A, RRID:AB_2621375
Anti-SUMO2/3	Abcam	Cat# ab3742, RRID:AB_304041
Anti-Lamin A/C	Santa Cruz	Cat#: sc-7292, RRID: AB_627875
Anti-BrdU	Singma-Aldrich	Cat#: B2531, RRID:AB_476793
Alexa Fluor 488 goat anti-mouse IgG	Thermo Fisher	Cat# A-11029; RRID:AB_2534088
Alexa Fluor 594 goat anti-mouse IgG	Thermo Fisher	Cat# A-11032; RRID:AB_2534091
Alexa Fluor 488 goat anti-rabbit IgG	Thermo Fisher	Cat# A-11034; RRID:AB_2576217
Alexa Fluor 594 goat anti-rabbit IgG	Thermo Fisher	Cat# A-11037; RRID:AB_2534095
Alexa Fluor 647 goat anti-mouse IgG	Thermo Fisher	Cat# A-21236, RRID: AB_2535805
Streptavidin-HRP	Abcam	Cat# ab7403
Streptavidin- Alexfluore488	Thermo Fisher	Cat# S32354
Anti-mouse IgG, HRP-linked Antibody	Cell Signaling	Cat# 7076S; RRID:AB_330924
Anti-rabbit IgG, HRP-linked Antibody	Cell Signaling	Cat# 7074S; RRID:AB_2099233
Bacterial and virus strains
NEB® 5-alpha Competent E. coli	NEB	Cat# C2987H
Laboratory lentivirus	This study	N/A
Chemicals, peptides, and recombinant proteins
Lipofectamine 2000	Thermo Fisher	Cat# 11668027
Lipofectamine RNAi Max	Thermo Fisher	Cat# 13778075
Dynabeads™ MyOne™ Streptavidin C1	Thermo Fisher	Cat# 65001
Biotin	Sigma-Aldrich	Cat# 58-85-5
A-485	Selleckchem	Cat# S7256
SGC-CBP30	Sigma-Aldrich	Cat# SML1133
Trichostatin A (TSA)	Calbiochem	Cat# 647925
TAK-981	Selleckchem	Cat# S8829
TelG-FITC	PNA bio	Cat# F1010
Dynabeads® Protein A	Thermo Fisher	Cat# 10002D
4-Hydroxytamoxifen (4-OH)	Sigma-Aldrich	Cat# H7904
Shield-1	Takara	Cat# 632189
FISH blocking reagent	Millipore Sigma	Cat# 11096176001
SuperScript™ III First-Strand Synthesis	Thermo Fisher	Cat# 11752050
Halt™ Protease Inhibitor Cocktail	Thermo Fisher	Cat# 78429
Fast SYBR™ green master mix	Thermo Fisher	Cat# 4385612
Critical commercial assays
Click-IT EdU	Thermo Fisher	Cat# C10637
Deposited data
List of proteins identified by BLOCK-ID	Table S1	ProteomeXchange: PXD057112
List of proteins identified by TRIM24 BioID	Table S2	ProteomeXchange: PXD063658
Original imaging data	This study	DOI: 10.17632/bw5v3pb3vy.1
Experimental models: Cell lines
Human: U2OS	ATCC	Cat#: HTB-96
Human: HeLa LT	Dr. Jan Karlseder (Salk Institute)	N/A
Human: LM216T	Dr. Roger Greenberg (University of Pennsylvania)	Verma et al.69
Human: LM216J	Dr. Roger Greenberg (University of Pennsylvania)	Verma et al.69
Human: SaOS2	ATCC	Cat#: HTB-85
Human: Cal72	Dr. Rachel Flynn (Boston University)	Flynn et al.70
Human: MG-63	ATCC	Cat#: CRL-1427
Human: SJSA1	ATCC	Cat#: CRL-2098
Human: HEK293T	ATCC	Cat#: CRL-3216
Human: U2OS shSCR	This study	N/A
Human: U2OS shTRIM24 (C1)	This study	N/A
Human: U2OS shTRIM24 (C2)	This study	N/A
Human: U2OS TurboID-TRF1	This study	N/A
Human: U2OS mCherry-TRF1-FOK1	Dr. Roger Greenberg (University of Pennsylvania)	Dilley et al.13
Human: U2OS mCherry-TRF1-FOK1 (D450A)	Dr. Roger Greenberg (University of Pennsylvania)	Dilley et al.13
Human: U2OS-LacO-I-SceI-TetO	Dr. Roger Greenberg (University of Pennsylvania)	Shanbhag et al.71
Human: U2OS PML KO	Dr. Eros Denchi (NCI)	Loe et al.20
Oligonucleotides
List of siRNAs	Supplementary Table S3	N/A
List of shRNAs	Supplementary Table S3	N/A
Recombinant DNA
pCW57.1 mycBirA*-LacI	This study	N/A
pcDNA5 FRT/TO mycBirA*-TRIM24	This study	N/A
pcDNA3.1 GFP-TRIM24	This study	N/A
pcDNA3.1 GFP-TRIM24 ΔRING	This study	N/A
pcDNA3.1 GFP-TRIM24 ΔPHD	This study	N/A
pcDNA3.1 GFP-TRIM24 ΔBRD	This study	N/A
pcDNA3.1 GFP-TRIM24 ΔCC	This study	N/A
pcDNA3.1 GFP-TRIM24 SD	This study	N/A
pcDNA5 FRT/TO Flag-TRIM24-TebDB	This study	N/A
pcDNA5 FRT/TO Flag-TebDB	This study	N/A
pcDNA3.1 HA-PML IV	This study	N/A
pcDNA5 FRT/TO GFP-MRE11	Miller Lab	N/A
pcDNA5 FRT/TO Myc-NBS1	Miller Lab	N/A
pcDNA5 FRT/TO GFP-TRIM24-TebDB	This Study	N/A
pcDNA5 FRT/TO Flag-TRIM24 ΔRING-TebDB	This study	N/A
pcDNA5 FRT/TO Flag-TRIM24 ΔCC-TebDB	This study	N/A
pcDNA5 FRT/TO Flag-TRIM24 ΔPHD-TebDB	This study	N/A
pcDNA5 FRT/TO Flag-TRIM24 ΔBRD-TebDB	This study	N/A
pcDNA5 FRT/TO Flag-TRIM24 SD-TebDB	This study	N/A
RMCE GFP-TRF1	Addgene	Addgene plasmid # 16464, RRID:Addgene_16464
pcDNA5 FRT/TO Flag-TRIM24	This study	N/A
pcDNA5 FRT/TO Flag-TRIM24 SD	This study	N/A
pcDNA5 FRT/TO Flag-TRIM24 ΔCC	This study	N/A
pcDNA5 FRT/TO Flag-TRIM24-TRF1	This study	N/A
pcDNA5 FRT/TO SFB	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-ATAD2	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-BAZ1A	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-BAZ1B	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-BAZ2A	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-BRD2	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-BRD4	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-BRD7	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-PHIP	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-SMARCA2	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-SMARCA4	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-TAF1	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-TRIM24	Kim et al.29	N/A
pcDNA5 FRT/TO SFB-ZMYND8	Kim et al.29	N/A
pcDNA3.1 HA-PML IV-TebDB	This study	N/A
pcDNA3.1 GFP-BLM	This study	N/A
pcDNA3.1 GFP-PIAS1	This study	N/A
pcDNA3.1 GFP-MMS21	This study	N/A
pcDNA5 FRT/TO Flag-TRF1	This study	N/A
pLKO.1 Scramble	Addgene	Addgene plasmid # 1864; RRID:Addgene_1864
pLKO.1 shTRIM24 (C1)	This study	N/A
pLKO.1 shTRIM24 (C2)	This study	N/A
pMD2.G	Addgene	Addgene plasmid # 12259; RRID:Addgene_12259
pPAX2	Addgene	Addgene plasmid # 12260; RRID:Addgene_12260
pOG44	Thermo Fisher	Cat# V6005-20
Software and algorithms
ImageJ (v.1.53)	Schneider et al.72	RRID:SCR_003070 https://imagej.nih.gov/ij/
Prism (v.9)	GraphPad	RRID:SCR_002798 https://www.graphpad.com
FV-10-ASW3.1	Olympus	https://www.olympus-ims.com/en/
FlowJo (V.10)	BD Bioscience	https://www.flowjo.com/solutions/flowjo

STAR METHODS

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

HEK293T (ATCC, CRL-3261, female), U2OS (ATCC, HTB-96, female), SaOS2 (ATTC, HTB-85, female), MG-63 (ATCC, CRL-1427, male), SJSA1 (ATCC, CRL-2098, male) were obtained from ATCC. WT/D450A TRF1-FokI U2OS (female), U2OS-LacO-I-SceI-TetO (female) cells were generously provided by Dr. Roger Greenberg (University of Pennsylvania). LM216T (male) and LM216J (male) were originally described by J.P. Murnane PMID2431271 and generously provided by Dr. Jan Karlseder (Salk Institute). HeLa LT (female) cell line was generously provided by Dr. Jan Karlseder (Salk Institute). Cal-72 (male) cell line was generously provided by Dr. Rachel Flynn (Boston University). All cells were tested negative for Mycoplasma using Mycoplasma PCR detection kit (ABM) and cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin, and maintained at 37°C in a 5% CO₂ atmosphere.

METHOD DETAILS

Antibodies

Antibodies used in this study are provided in Supplemental Table 4.

Plasmids

To generate mycBirA*-LacI, the LacI sequence was amplified using Q5 DNA polymerase (NEB) from the mCherry-LacI vector²⁸. The PCR product was inserted into the mycBirA* pcDNA3.1 vector²³ using the XhoI and AflII restriction enzyme sites. This construct was then subcloned into the pCW57.1 vector using the NheI and EcoRI restriction enzyme sites. mycBirA* was amplified from the mycBirA* pcDNA3.1 vector and inserted into pcDNA5 TO vector using the AflII and Xho1 enzyme sites (mycBirA* pcDNA5 FRT/TO). TRIM24 was amplified using Q5 DNA polymerase from Flag-TRIM24 WT (Miller lab) and inserted into the mycBirA* pcDNA5 FRT/TO using the BamHI and XhoI sites. Functional domains of TRIM24 were identified using the UniProt/Prosite database. Deletion mutants of TRIM24 (ΔRING, ΔCC, ΔPHD, and ΔBRD) were generated by PCR amplification and inserted into custom-built expression vectors with N-terminal eGFP (GFP-pcDNA3.1, Miller lab) or Flag (Flag-pcDNA5 FRT/TO, Miller lab) tags. SUMO-deficient TRIM24 mutants (K723R and K741R) were generated via site-directed mutagenesis and inserted into both GFP- and Flag-pcDNA5 FRT/TO vectors using the XhoI and AlfII restriction enzyme sites. The DNA binding domain of Teb1 was synthesized and C-terminally fused to TRIM24 and PML IV through overlap extension PCR. The resulting PCR products were inserted into GFP-pcDNA3.1 and Flag-pcDNA5 FRT/TO vectors using the XhoI and AlfII restriction enzyme sites. TRIM24 shRNAs (C1 and C2) were cloned into the pLKO.1 TRC vector according to the protocols from Addgene (https://www.addgene.org/protocols/plko/).

Cell line generation

mycBirA*-LacI construct cloned into the pCW57.1 vector and shRNAs cloned into the pLKO.1 vector were introduced into target cells via lentiviral transduction. Briefly, mycBirA*-LacI in pCW57.1 or shRNAs in pLKO.1, along with the packaging vectors pPAX2 and pMD2.G, were co-transfected into HEK293T cells using Lipofectamine 2000 (ThermoFisher) according to the manufacturer’s instructions. Lentivirus particles were harvested at 48 and 72 h post-transfection, centrifuged at 500 ×g for 10 min, and filtered through 0.45 μm PES syringe filters to remove cell debris. The virus-containing media were then applied to the target cells. The following day, the lentivirus media was replaced with fresh culture media containing the indicated antibiotics for selection. Detailed information on the shRNAs used in this study is provided in Table S3. mycBirA*-TRIM24 construct was cloned into the pcDNA5 FRT/TO vector and along with pOG44, were transfected into U2OS FRT cells using Lipofectamine 2000 (ThermoFisher). The following day, the media was replaced with fresh media containing antibiotics for selection. Expression of mycBirA*-TRIM24 was validated by immunofluorescence and Western blotting.

Clonogenic assays

siRNAs were introduced using RNAiMAX (ThermoFisher) following the manufacturer’s recommendations at a final concentration of 10 nM. Next day, cells were trypsinized and seeded at a density of 500 cells per well in triplicate in six-well plates. Cells were incubated for 10 to 14 days until colonies formed. Colonies were stained with 0.5% crystal violet in 20% ethanol and manually counted. The efficiency of siRNA-mediated knockdown was assessed by Western blot or qPCR. Detailed information on the siRNAs used in this study is provided in Table S3.

Confocal microscopy and Immunofluorescence

Cells were plated on glass coverslips and subjected to the desired experimental conditions. Prior to fixation, a pre-extraction step was performed by incubating the cells on ice do 5 min in CSK buffer containing 10 mM PIPES, pH 6.8, 100 mM NaCl, 300 mM sucrose, 3 mM MgCl₂, 1 mM EGTA and incubated on ice for 5 min. The cells were then washed three times with cold PBS and fixed with 3% paraformaldehyde (PFA). Following fixation, cells were permeabilized with 0.4% Triton X-100 for 15 minutes at room temperature. Blocking was performed for 1 h at room temperature using a buffer containing 3% BSA and 0.4% Triton X-100. Then, primary antibodies, diluted in the blocking buffer, were applied and incubated overnight at 4°C. The next day, Alexa Fluor-conjugated secondary antibodies (ThermoFisher) were applied and incubated for 1 h at room temperature. Then, coverslips were mounted onto slides with Vectashield mounting medium containing DAPI (Vector Laboratories). To detect cells in S-phase, cells were incubated with 10μM EdU for 30 min. EdU was labeled using the Click-iT EdU imaging Kit (Thermo Fisher) as instructed by the manufacturer. After EdU labeling, subsequent immunofluorescence steps were performed as described above. Immunofluorescence signals were visualized using an inverted FV3000 scanning confocal microscope (Olympus) with Olympus PlanApo N 60x/1.40 Oil 0.17/FN22 Microscope Objective (Olympus), controlled by FW31S software (Olympus). Detailed information on the antibodies used in this study is provided in Table S4.

Immunoprecipitation and Western blotting

2 × 10⁶ cells were lysed using 1 ml of NETN buffer supplemented with 1X HALT protease inhibitor (Thermo Fisher) and 1X HALT phosphatase inhibitor (Thermo Fisher). To detect SUMOylation of TRIM24 and TRIM24-TebDB, cells were lysed using 1 ml of RIPA buffer (50 mM TRIS pH 7.0, 150 mM NaCl, 1 % NP-40, 1 % Sodium deoxycholate, 0.1 % SDS) containing 20 mM N-ethylmaleimide (Sigma-Aldrich). After incubation with 1 unit of Turbonuclease (Accelagen), insoluble components were removed by centrifugation at 16,500 × g for 10 min at 4°C. The supernatant was collected and incubated with 1 μg of the desired antibodies and 20 μl of protein A beads (Thermo Fisher) overnight at 4°C. The following day, beads were washed three times with the lysis buffer and then two times with 50 mM TRIS. Bound proteins were eluted from the beads by adding 2X SDS-PAGE sample buffer. The eluates were boiled at 95°C for 5 min, separated by SDS-PAGE, and transferred onto a nitrocellulose membrane (GE Healthcare) for 90 min at 110 V. After blocking with 5% BSA in TBS-T (0.05% Tween-20), primary antibodies, diluted in the blocking buffer, were applied overnight at 4°C. The membrane was washed three times with TBS-T for 10 min each at room temperature and then incubated with HRP-conjugated secondary antibodies (Cell Signaling Technologies) for 1 h at RT. Chemiluminescence signals, following application of ECL (GE Healthcare), were detected using Chemidoc MP Imaging System (Bio-Rad).

SIRF assay

To examine the localization of proteins at stalled replication forks, cells on coverslips were incubated with 125 mM of EdU for 8 min and then treated with 4 mM HU for 3 h. After fixation with 3 % PFA, EdU was conjugated to biotin azide (Thermo Fisher) by incubating click reaction buffer containing 1X PBS, 2mM copper sulfate, 10 μM biotin azide (Thermo Fisher) and 100 mM sodium ascorbate. Subsequently, primary antibodies against biotin and either TRIM24 or BLM were applied and incubated at 4°C overnight. The next day, PLA was performed using Duolink^® PLA kit (Sigma-Aldrich) according to the manufacturer’s instruction. Briefly, after overnight incubation with the primary antibodies, In Situ PLA probes (anti-mouse and anti-rabbit) were applied to coverslips for 1 h at 37 4°C. The PLA ligase mix was then added and incubated for 30 min at 37°C. Following three washes, the PLA amplification mix was applied and incubated for 90 min at 37°C. After additional washes, the coverslips were mounted on slides using Vectashield mounting medium containing DAPI. SIRF foci were visualized using an inverted FV3000 scanning confocal microscope (Olympus) with Olympus PlanApo N 60x/1.40 Oil 0.17/FN22 Microscope Objective (Olympus), controlled by FW31S software (Olympus). Detailed information on the antibodies used in this study is provided in Table S4.

qPCR

Total RNA was extracted from cells using the RNeasy kit (Qiagen) as instructed by the manufacturer. cDNA was generated from the extracted RNA using Superscript III First-Strand Synthesis kit (Thermo Fisher). Quantitative PCR was conducted on the StepOnePlus Real-time PCR system (Applied Biosystems) with Fast SYBR^™ green master mix (Thermo Fisher). Detailed information on the qPCR primers in this study is provided in Table S5.

Proliferation assays

siRNAs were introduced using RNAiMAX (ThermoFisher) following the manufacturer’s recommendations at a final concentration of 10 nM. After 24 h post-transfection, 9 × 10³ cells were seeded in 12-well plates. Proliferation of the cells was monitored daily from day 1 to day 5. Growth curve was quantified by crystal violet staining. Briefly, cells were fixed and stained with crystal violet, and excess dye was removed by washing the wells three times with distilled water. The bound dye was solubilized with 10% SDS, and absorbance at 590 nm was measured using a Synergy H1 microplate reader (BioTek). Representative images of the stained cells were acquired using Chemidoc MP Imaging System (Bio-Rad).

Detection of ssTelo foci

Detection of ssTelo foci was carried out as previously described²⁰. After completing the immunofluorescence procedure as described earlier, cells were re-fixed with 3% PFA. Then, the coverslips were incubated with the blocking buffer containing 500 μg/mL Rnase A (Thermo Fisher) for 1 h at 37°C. After Rnase A treatment, the coverslips were dehydrated through a series of ethanol washes (70%, 90%, and 100%). Once air-dried, FITC-OO[TTAGGG]3-labeled PNA probe (PNA Bio), diluted in 10 mM Tris pH 7.4, 70% formaldehyde, 1 mg/ml FISH blocking reagent (Roche), were applied and incubated with the coverslips at room temperature for 3 h. The coverslips were then washed twice with 70% formaldehyde in PBS, followed by two additional washed with 1X PBS. The coverslips were mounted onto slides with Vectashield mounting medium containing DAPI.

Cell cycle analysis

Cell cycle analysis was performed as previously described, with minor modifications⁶³. Briefly, cells were incubated in medium containing 10 μM BrdU for 30 min prior to collection. After trypsinization, cells were washed twice with PBS and fixed in 70% ethanol for 30 min on ice. Fixed cells were then permeabilized using a buffer containing 2M HCl and 0.5% Triton X-100 in PBS for 30 min at room temperature on a rocking platform. Next, cells were incubated with 0.1 M Na₂B₄O₇ (pH 8.5) and washed with 1% BSA/PBS containing 0.3 mM EDTA. Cells were then incubated with an anti-BrdU antibody (Sigma-Aldrich) for 1 h, followed by Alexa Fluor-conjugated secondary antibodies (Thermo Fisher) for 45 min at RT. After washing again with 1% BSA/PBS containing 0.3 mM EDTA, cells were stained with 4 μg/ml propidium iodide and treated with 100 μg/ml RNase A (Thermo Fisher) for 30 min. Samples were analyzed by BD LSRFortessa SORP Flow Cytometer (BD Biosciences), and data were processed using FlowJo software.

Telomere DNA synthesis detection by EdU

For the siRNA-mediated knockdown conditions, 5 pM of siRNAs was introduced to 800,000 cells using Dharmafect (Horizon Discovery). 48 h later, cells were transfected with a FLAG-tagged TRF1-FokI expressing plasmid. Cells were pulsed with 10μM of EdU for 1 h before fixation. Cells on coverslips were washed with PBS and fixed by incubating with 2% PFA for 10 min. Cells were permeabilized with 0.1% (w/v) sodium citrate and 0.1 % (v/v) Triton X-100 for 5 min and then washed with PBS before incubating with blocking solution (1mg/mL BSA, 10% normal goat serum, 0.1% Tween) for 30 min. Primary antibodies were diluted in blocking solution and added to cells overnight at 4°C. Next, cells were washed three times with PBS for 5 min and incubated with Alexa Fluor-conjugated secondary antibodies (Life Technologies) for 1h at room temperature. After incubation, cells were washed three times, fixed again in 2% PFA for 10 min followed by two additional PBS washes prior to performing click-reaction using the Click-IT Plus EdU Cell Proliferation Kit (Invitrogen) to detect EdU. For experiments with A-485 and SGC-CBP30, U2OS cells expressing mCherry-TRF1-FokI were cultured with inhibitors at the indicated concentrations for 3 days. 24 h prior to harvest, cells were induced with doxycycline (40ng/μL1) followed by the addition of 1μM of tamoxifen (4-OHT, Sigma) and 1μM of Shield1 Ligand (Takara Clontech) for 3 h before harvest. Cells were pulsed with EdU (10μM) for 1 h before processing as above. For the complementation experiments, 48 h post knockdown using siCTRL or TRIM24 3’UTR siRNA, cells were transfected for 6 h with mCherry-tagged TRF1-FokI expressing plasmid alone or in combination with FLAG-tagged variants of TRIM24. Next day, cells were pulsed with EdU (10μM) for 1hr and then harvested for IF as described above.

Telomere restriction fragment analysis by pulsed-field gel Electrophoresis (PFGE)

Telomere gels were performed using telomere restriction fragment (TRF) analysis. Genomic DNA was extracted from the different cell lines and digested overnight with AluI and MboI restriction enzymes (NEB). Digested DNA was purified and 2–5 μg of DNA was run on a 1% PFGE agarose gel (Bio-Rad) in 0.5 × TBE buffer using the CHEF-DRII system (Bio-Rad) at 6V cm⁻¹; initial switch time 1 s, final switch time 6 s, for 17 h at 14°C. The gel was then dried for 2 h at 60°C, denatured in a 0.5 N NaOH 1.5 M NaCl solution, and neutralized. Gel was hybridized with ³²P-labeled (TTAGGG)₄ oligonucleotides overnight at 55°C in UltaHyb Hybridization Buffer (ThermoFisher Catalog number: AM8669). The next day, the membrane was washed three times in 2x SSC and once in 2x SSC supplemented with 0.5% SDS, exposed onto a phosphor screen, and scanned using Typhoon 9400 PhosphoImager (GE Healthcare).

Chromosome Orientation FISH (CO-FISH)

CO-FISH was performed as described⁶⁴. In brief, cell cultures were incubated with 7.5mM BrdU and 2.5mM BrdC for ~14 h. After removal of nucleotide analogs, colcemid (GIBCO) was added for ~2h and cells were harvested by trypsinization, swelled in 0.075M KCl for approximately 7 min at 37°C and fixed in cold fixative (70% Methanol: 30% Glacial Acetic Acid). Metaphase chromosomes were spread by dropping them onto slides, then RNase A (0.5 mg/mL) was treated. Slides were incubated in 2 × SSC containing 0.5 mg/mL Hoechst 33258 for 15 mins in the dark and irradiated for 3 min (5.4 × 10⁵ J/m², energy 5400) in a UV Stratalinker 2400 (Stratagene). The nicked BrdU/C substituted DNA strands are degraded by Exonuclease III digestion for 1 h. The slides were then washed in PBS, dehydrated by ethanol washes and allowed to air dry completely. The remaining strands were hybridized with fluorescence labeled DNA probes. The positive telomere strand (polymerized by lagging strand synthesis) was labeled using Alexa Fluor 488-conjugated (TTAGGG)₄, and the negative telomere strand (polymerized by leading strand synthesis) was labeled using Alexa Fluor 568 conjugated (CCCTAA)₄. Prior to hybridization of the first PNA, DNA is denatured by heating at 70°C for 10 min, and then incubated for 2 h at room temperature. Slides were washed for 15 min with Wash Solution A (70% Formamide and 10mM Tris-HCl pH 7.0–7.5, BSA), dried, and then incubated with the second PNA for 2 h at room temperature. The slides were then washed again twice for 15 min with Wash Solution A and 3 times with Wash Solution B (0.1M Tris-HCl pH7.2, 0.15M NaCl and 0.08% Tween) for 5 min at room temperature. Finally, cells were dehydrated by ethanol washing as mentioned above and mounted using ProLong Gold Mounting Media with DAPI. Metaphase chromosomes were visualized by conventional florescence microscope with a 63X Plan λ objective (1.4 oil) on a Nikon 90i microscope.

Pulsed-field gel Electrophoresis (PFGE)

Telomere gels were performed using telomere restriction fragment (TRF) analysis. Genomic DNA was extracted from the different cell lines and digested overnight with AluI and MboI restriction enzymes (NEB). Digested DNA was purified and 2–5 μg of DNA was run on a 1% PFGE agarose gel (Bio-Rad) in 0.5 × TBE buffer using the CHEF-DRII system (Bio-Rad) at 6V cm⁻¹; initial switch time 1 s, final switch time 6 s, for 17 h at 14°C. The gel was then dried for 2 h at 60°C, denatured in a 0.5 N NaOH 1.5 M NaCl solution, and neutralized. Gel was hybridized with ³²P-labeled (TTAGGG)₄ oligonucleotides overnight at 55°C in UltaHyb Hybridization Buffer (ThermoFisher). The next day, the membrane was washed three times in 2x SSC and once in 2x SSC supplemented with 0.5% SDS, exposed onto a phosphor screen, and scanned using Typhoon 9400 PhosphoImager (GE Healthcare).

BioID pull-down and on-bead digestion

BioID pull-down was performed as previously described^23,65. Briefly, biotinylation in living cells was achieved by adding 50 μM biotin to the culture media and incubating overnight. The following day, cells from each experimental set were counted and lysed using BioID lysis buffer (50 mM Tris, pH 7.4, 500 mM NaCl, 0.4% SDS, 1 mM dithiothreitol, and 1X HALT protease inhibitor). For the BLOCK-ID, 2 × 10⁸ cells were used while for BioID analysis of TRIM24, 4 × 10⁷ cells were used. After lysis, the cell suspension was subjected to two rounds of tip sonication at power 1 for 1 min each (20 seconds on, 20 seconds off) using a FB705 ThermoFisher sonicator. Insoluble cellular debris was removed by centrifugation at 16,500 ×g for 10 min at 4°C. The supernatant was then incubated with Dynabeads MyOne C1 streptavidin beads (ThermoFisher) overnight at 4°C with gentle rotation. After incubation, the beads were collected using a magnetic stand and sequentially washed once with Wash buffer 1 (2% SDS), once with Wash buffer 2 (0.1% deoxycholate, 1% Triton X-100, 500 mM NaCl, 1 mM EDTA, and 50 mM HEPES), once with Wash buffer 3 (250 mM LiCl, 0.5% NP-40, 0.5% deoxycholate, 1 mM EDTA, and 10 mM Tris, pH 7.4), and twice with 50 mM Tris. The captured proteins were subjected to on-bead digestion as previously described⁶⁶. Briefly, the beads were resuspended in a solution containing 50mM triethylammonium bicarbonate (Sigma-Aldrich) and 2 M urea. Reducing of proteins was performed by incubating the beads with 20 mM DTT for 30 min, followed by alkylation with 20 mM iodoacetamide for 30 min in the dark. Then, 1 μg of sequencing grade trypsin (Promega) was added to the beads, and the mixture was incubated at 37°C overnight on a shaker. The following day, the digested peptides were collected and cleaned using a custom-built reversed-phase C18 column.

Liquid chromatography-mass spectrometry

The LC-MS/MS analysis was conducted as described in these previous publications with minor modifications^67,68. The peptides were sequenced using the Orbitrap Fusion Lumos Tribrid Mass Spectrometer and the Ultimate3000 RSLCnano liquid chromatography system (ThermoFisher). The peptides, which were reconstituted in 15 μl of 0.1% formic acid, were injected into an Acclaim PepMap100 Nano-Trap Column (100 μm × 2 cm, ThermoFisher Scientific, San Jose, CA, USA) containing 5 μm C18 particles at a flow rate of 5 μl per min. A linear gradient of 8% to 28% solvent B (0.1% formic acid in 95% acetonitrile) was used to separate the peptides at a flow rate of 300 nl/min over 95 min on an EASY-Spray column (50 cm × 75 μm ID, Thermo Fisher) packed with PepMap RSLC C18 and 2 μm C18 particles (Thermo Fisher). An EASY-Spray ion source was used for ionization at 2.4 kV. The mass spectrometry analysis was conducted in a data-dependent mode with a full scan in the mass-to-charge ratio (m/z) range of 350 to 1800 in the “Top Speed” setting, with a three-second cycle. MS1 for the precursor ions and MS2 for the fragmentation ions were acquired at a resolution of 120,000 and 30,000 at an m/z of 200, respectively. The MS2 fragmentation was carried out using the higher-energy collisional dissociation (HCD) method at 32 of normalized collision energy (NCE) with 5% stepped NCE. The automatic gain controls were set to 1 million ions for MS1 and 0.05 million ions for MS2, and the maximum ion injection time was set to 50 milliseconds for MS1 and 100 milliseconds for MS2. The precursor isolation window was set to 1.6 m/z with a 0.4 m/z of offset, and dynamic exclusion was set to 30 seconds while singly charged ions were rejected. Calibration was carried out using the lock mass option (m/z 445.1200025) from ambient air.

Database search

The database search was conducted as described in these previous publications with minor modifications^67,68. To identify and quantify proteins, the acquired spectra were searched against the human UniProt database (released in December 2019, containing protein entries with common contaminants) using the SEQUEST search algorithm embedded in the Thermo Proteome Discoverer platform (version 2.2, ThermoFisher Scientific). For MS/MS preprocessing, the top 10 peaks in each 100 m/z window were selected for the database searches. The following search parameters were employed: a) trypsin was set as a proteolytic enzyme (with up to two missed cleavages); b) 20 ppm was set as peptide mass error tolerance; c) 0.02 Da was set as fragment mass error tolerance; d) carbamidomethylation of cysteine (+57.02146 Da) was set as fixed modification; e) oxidation of methionine (+15.99492 Da) and protein acetylation (+ 42.01056 Da) on N-terminus were set as dynamic modifications; and f) the minimum peptide length was set to 6 amino acids. Peptides and proteins were filtered at a 1 % false-discovery rate (FDR) at the PSM level using the percolator node and at the protein level using the protein FDR validator node, respectively.

Analysis of mass-spectrometry result

After mass-spectrometry analysis, proteins identified by the BirA* control were excluded. Additionally, unique proteins found in the proliferating condition and those enriched more than 1.5-fold compared to the growth-arrested condition were included. For the BLOCK-ID analysis, Proteins exclusively identified in the proliferating condition, as well as those exhibiting a fold enrichment of greater than 1.5 times (based psm) are presented. The fold enrichment representing the ratio of protein abundance in proliferating versus growth-arrested conditions (normal/arrested) is expressed as Log2 value. For TRIM24 BioID, those exhibiting a fold enrichment of greater than 1.5 times (based psm) are presented. To determine the relative abundance of each protein, the total PSM for each protein was normalized to account for the total length of the protein in amino acids. The relative abundance of each protein finally was expressed as percentage of the sum of all adjusted PSM except those from BirA*-TRIM24.

Quantification and Statistical Analysis

All data in this study were analyzed in GraphPad Prism (v.10.2.0), ImageJ (v.1.53a), and Microsoft Excel. No statistical methods were used to predetermine the sample size. Two-tailed Student’s t-test, one-way analysis of variance by a Dunnett multiple comparison test, one-way Anova and Mann-Whitney test were used to determine statistical significance as indicated in each figure legend. P-values are specified in the figures. Unless otherwise mentioned, NS indicates Non-significant. Dots on the graphs indicates number of biologically independent experiments, indicated by the n in the figure legends and indicated number represents total number of cells or treatments analyzed per each experiment.

Supplementary Material

2

Table S1. List of proteins identified by BLOCK-ID mass-spectrometry analysis, related to Figure 1.

3

Table S2. List of proteins identified by TRIM24 mass-spectrometry analysis, related to STAR Methods.

Highlights.

• BLOCK-ID allows interrogation of proteins at stressed replication forks

• TRIM24 is a mediator of ALT telomere extension

• A p300/CBP-SUMO chromatin signaling pathway regulates TRIM24 function at telomeres

• Tethering TRIM24 to telomeres bypasses the need for PML in telomere DNA synthesis