Abstract The CHAMP1 complex, a little-known but highly conserved protein complex consisting of CHAMP1, POGZ, and HP1α, is enriched in heterochromatin though its cellular function in these regions of the genome remain unknown. Here we show that the CHAMP complex promotes heterochromatin assembly at multiple chromosomal sites, including centromeres and telomeres, and promotes homology-directed repair (HDR) of DNA double strand breaks (DSBs) in these regions. The CHAMP1 complex is also required for heterochromatin assembly and DSB repair in highly-specialized chromosomal regions, such as the highly-compacted telomeres of ALT (Alternative Lengthening of Telomeres) positive tumor cells. Moreover, the CHAMP1 complex binds and recruits the writer methyltransferase SETDB1 to heterochromatin regions of the genome and is required for efficient DSB repair at these sites. Importantly, peripheral blood lymphocytes from individuals with CHAMP1 syndrome, an inherited neurologic disorder resulting from heterozygous mutations in CHAMP1, also exhibit defective heterochromatin clustering and defective repair of DSBs, suggesting that a defect in DNA repair underlies this syndrome. Taken together, the CHAMP1 complex has a specific role in heterochromatin assembly and the enhancement of HDR in heterochromatin. Subject terms: Homologous recombination, Double-strand DNA breaks, Telomeres, Histone post-translational modifications __________________________________________________________________ The CHAMP1 complex, consisting of CHAMP1, POGZ, and HP1α, is enriched in heterochromatin, but its role was unclear. Here, the authors demonstrate that the CHAMP1 complex promotes heterochromatin assembly and homology-directed repair, with its dysfunction linked to CHAMP1 syndrome. Introduction DNA double-strand breaks (DSBs), resulting from intrinsic DNA replication errors or from external exposure to DNA damaging agents, can occur at any site throughout the genome. Homology-directed repair (HDR), including classical homologous recombination (HR) and break-induced replication (BIR), functions to repair these DSBs. Some regions of chromosomes may require specific mechanisms of DSB repair. For example, heterochromatic regions of chromosomes, such as centromeres and telomeres, are highly-compacted structures and pose a significant barrier to HDR-mediated DSB repair. Accordingly, cells may require distinct mechanisms to mediate HDR in these chromosomal regions^[40]1–[41]4. Heterochromatin regions of the genome are characterized by increased trimethylation of histone H3 on lysine 9 (H3K9me3), a key modification that interacts with histone reader proteins^[42]5 and is critical for heterochromatin clustering and function^[43]6. Heterochromatin assembly is regulated by the coordinated activity of writer H3K9 methyltransferases and eraser H3K9 demethylases. These enzymes regulate the level of H3K9me3 in heterochromatin^[44]6,[45]7, and these mechanisms are coupled to the HDR repair process^[46]8–[47]12. Some heterochromatin structures have a unique nucleosome organization, and these structures may also require specialized mechanisms for heterochromatin assembly and local HDR repair. For example, ~10–15% of cancers employ a telomerase-independent mechanism known as the alternative lengthening of telomeres (ALT pathway). This process relies on HDR for template-driven telomere lengthening by a BIR-like mechanism^[48]13,[49]14. Moreover, ALT tumor cells have a high level of H3K9me3 and a specialized heterochromatic microenvironment that facilitates DNA synthesis and telomere expansion^[50]15,[51]16. Whether the CHAMP1 complex plays a specific role in ALT telomere maintenance or ALT telomere DSB repair remains unknown. A previous report^[52]17 identified a distinct heterochromatin complex- namely, the CHAMP1/POGZ/HP1α complex- with H3K9me3-binding activity. CHAMP1 (CHromosome Alignment Maintaining Phosphoprotein 1), a highly-conserved zinc-finger protein, functions as a regulator of chromosome segregation in mitosis and is required for the correct alignment of chromosomes on the metaphase plate^[53]18. The CHAMP1 binding partner POGZ (POGO transposable element with ZNF domain), also a zinc-finger protein, functions in mitotic cell cycle progression and kinetochore assembly^[54]19. HP1α, a member of the HP1 family of “hub proteins”, interacts with several chromosomal proteins and binds to H3K9me3 through its chromodomain (CD). Whether this CHAMP1/POGZ/HP1α complex regulates heterochromatin assembly or mediates HDR in condensed regions of chromosomes remains unknown. Humans with de novo heterozygous mutations in CHAMP1, known as CHAMP1 syndrome, present with a neurodevelopmental disorder characterized by intellectual disability, behavioral symptoms, and distinct dysmorphic features^[55]20–[56]24. Interestingly, de novo heterozygous mutations in the POGZ gene also result in a rare but highly-related neurodevelopment syndrome called the White Sutton Syndrome (WSS)^[57]25–[58]33. Whether these inherited mutations result from defects in heterochromatin assembly or HDR in heterochromatin remains unknown. Recent studies have elucidated at least one mechanism by which CHAMP1 and POGZ promote HR^[59]34–[60]36. CHAMP1 has a unique, high-affinity site for the DNA repair HORMA protein, REV7. By binding to REV7, the CHAMP1/POGZ complex steals REV7 from the Shieldin (REV7/SHLD) complex^[61]36. The Shieldin complex normally blocks DNA end resection, promotes NHEJ repair, and inhibits HR repair^[62]37–[63]41. By binding to REV7, the CHAMP1/POGZ complex increases DSB end resection and promotes HR repair. Whether this mechanism of enhanced HR repair occurs in heterochromatin or whether its use is confined to euchromatic regions remains unknown. In the current study, we show that the CHAMP1/POGZ/HP1α heterochromatin complex performs a direct role in heterochromatin clustering and HDR repair. The complex promotes the deposition of H3K9me3, the clustering of heterochromatin, and the recruitment of HDR proteins (both HR proteins and BIR proteins) to DSBs at heterochromatic sites. While the complex subunits promote heterochromatin clustering at multiple sites, including centromeres and telomeres, it has a distinct and obligate role in maintaining ALT telomeres. Depletion of the CHAMP1/POGZ/HP1α complex results in reduced H3K9me3 and loss of heterochromatin clustering. Interestingly, the POGZ subunit also binds and recruits the methyltransferase SETDB1 to sites of heterochromatin. Finally, peripheral blood lymphocytes from individuals with CHAMP1 syndrome also exhibit defective heterochromatin clustering and HR, suggesting that a defect in DNA repair underlies this syndrome. Taken together, the CHAMP1/POGZ/HP1α is a specific molecular machine that regulates heterochromatin clustering and contributes to local HDR repair in heterochromatin. Results In silico protein-protein interaction modeling of the heterochromatic CHAMP1-POGZ-HP1α complex Quantitative proteomic analyses of histone epigenetic marks and readers identified CHAMP1-POGZ-HP1α as a heterochromatin complex that binds H3K9me3^[64]17. Recent studies have shown that the CHAMP1^[65]34,[66]36 and POGZ subunits^[67]35 promote HR-mediated DSB repair through interaction with REV7. Nevertheless, the relationship between the heterochromatin localization of the complex and HR is unclear. To investigate the organization of the multi-subunit CHAMP1 complex and design separation-of-interaction point mutants and/or domain-deletion mutants for the subunits of the complex, we performed in silico protein-protein interaction modeling with AlphaFold2/3-Multimer (AF)^[68]42,[69]43. This approach enabled us to predict a high-confidence interaction between the CHAMP1 N-terminal ZnF domain (1–87 aa) and the structured C-terminus of POGZ (1021–1410 aa) (Fig. [70]1a, b). The interaction between CHAMP1 and POGZ is formed primarily by hydrophobic residues at the interface (Fig. [71]1a). Supporting this model, N-terminal mutants of CHAMP1 (∆N, ∆N∆C and QLT/RRR) failed to bind to POGZ (Fig. [72]1c, [73]d, lanes 7, 8, and Supplementary Fig. [74]1a, [75]b lane 6). Furthermore, we also obtained a high-confidence structural model of POGZ and HP1α interaction (Supplementary Fig. [76]1c, d). Given that HP1α homodimerizes through its C-terminal chromo shadow domains based on the published crystal structure (PDB:3I3C) and POGZ is shown to interact with HP1α using its zinc-finger-like HP1-binding motif (HPZ)^[77]19, we used dimeric HP1α to predict the structural interface between two molecules of POGZ-HPZ motif and HP1α dimer. Our structural model indicates the key roles zinc-finger forming residues in the POGZ-HPZ motif play in maintaining the interface between POGZ and HP1α supporting previous mutational analyses^[78]19 (Supplementary Fig. [79]1c, d). The HP1α homodimerization interface in the predicted model matches the reported crystal structure (PDB:3I3C). We also investigated whether there is a direct interaction between CHAMP1 and HP1α. Our AF prediction of CHAMP1 and HP1α interface resulted in a low-confidence structural model. Despite the low confidence in the prediction, our immunoprecipitation (IP) data indicated that CHAMP1 can interact with HP1α independently of POGZ indicating that there is either a direct, yet unknown, interface between CHAMP1 and HP1α or an indirect interaction through an unknown protein (Fig. [80]1d, lane 7). For this interaction, our IP data indicate that the CHAMP1 C-terminal (Fig. [81]1d, lane 8, and Fig. [82]1e, lane 10) and dimerization^[83]19,[84]44 of HP1α is essential (Fig. [85]1f, lane 6). Conversely, HP1α is previously shown to interact with H3K9me3 using its N-terminal chromo domain^[86]45. Fig. 1. CHAMP1-POGZ-HP1α form a REV7-independent heterochromatin complex. [87]Fig. 1 [88]Open in a new tab a AlphaFold2-Multimer (AF2)-predicted structural model of POGZ_Cter (peach)-CHAMP1_NZnF (cyan) complex. The key residues (shown in sticks) contributing to the protein-protein interactions are highlighted in the inset including CHAMP1 residues (Q56, L61, and T70) that are mutated to disrupt the interaction between POGZ and CHAMP1. b The AF2-predicted model of POGZ_Cter-CHAMP1_NZnF is colored based on the confidence in the model prediction (100-high (blue) and 50-low (red)) and corresponding predicted aligned error plot (PAE) matrix showing the confidence in the predicted interaction between CHAMP1_NZnF and POGZ_Cter domains along with interface-predicted template scores (iPTM) for the AF2 prediction. c (Top) Schematic of CHAMP1 wild-type (WT) and mutants. N-ZNFs (N-terminus C2H2-Zn finger domains), SPE (PxxSPExxK motifs), WK (SPxxWKxxP motifs), FPE (FPExxK motifs), and C-ZNFs (C-terminus C2H2-Zn finger domains). 2A indicates W334A/K335A mutant, which loses REV7 interaction. (Bottom) Schematic of CHAMP1 complex connecting two or multiple H3K9me3 sites. The N-terminus of CHAMP1 binds to the C-terminus of POGZ. Additionally, both the C-terminus of CHAMP and the HPZ motif of POGZ can independently bind to HP1α and H3K9me3. d Western blot showing GFP-immunoprecipitation of GFP-empty vector, GFP-CHAMP1 WT or N- or C- terminal deletion mutants, and the co-immunoprecipitation of endogenous POGZ, HP1α and REV7. e Western blot showing GFP-immunoprecipitation of GFP-empty vector, GFP-CHAMP1 WT, and GFP-CHAMP1 mutants in 293 T cells, and the co-immunoprecipitation of endogenous POGZ, HP1α and REV7. f Western blot showing GFP-immunoprecipitation of GFP-empty vector, GFP-HP1α WT or I165E mutant, and the co-immunoprecipitation of endogenous POGZ, CHAMP1 and H3K9me3. GAPDH acts as a negative control. g Western blot showing GFP-immunoprecipitation of GFP-empty vector, GFP-HP1α in 293 T WT and sgCHAMP1 cells, and the co-immunoprecipitation of endogenous POGZ, CHAMP1, REV7, and H3K9me3. GAPDH acts as a negative control for IP. h Western blot showing GFP-immunoprecipitation of GFP-empty vector, GFP-REV7 in 293 T WT and sgCHAMP1 cells, and the co-immunoprecipitation of endogenous POGZ, CHAMP1, HP1α, and H3K9me3. GAPDH acts as a negative control for IP. All the immunoblots are representative of at least two independent experiments. Source data are provided as a Source Data file. Through these AF2 predictions, pull-down data, and previously published studies^[89]5,[90]19,[91]44,[92]46, we made a protein interaction map of the CHAMP1 complex containing CHAMP1 and POGZ (Fig. [93]1c). In this map, the N-terminus of CHAMP1 interacts with POGZ. Our pull-down data with multiple mutant proteins provides additional information regarding the interactions of this complex with HP1α and H3K9me3. Specifically, the C-terminus of CHAMP1 appears to directly/indirectly engage with an HP1α dimer^[94]5. Moreover, POGZ appears to interact with the HP1α dimer through its HPZ motif. The multivalency of the CHAMP1-POGZ-HP1α complex may form a cluster of interactions involving H3K9me3. Interestingly, the GFP-CHAMP1-2A mutant, which failed to bind to REV7, still pulled down POGZ and HP1α (Fig. [95]1e, lane 8), suggesting that REV7 is not required for the formation of the CHAMP1/POGZ/HP1α heterochromatin complex. GFP-HP1α successfully pulled down CHAMP1, POGZ, and H3K9me3, but did not pull down REV7 (Fig. [96]1g, lanes 5, 6). Moreover, GFP-REV7 pulled down CHAMP1 and POGZ but did not pull down HP1α or H3K9me3 (Fig. [97]1h, lanes 5, 6). There is no overlap between REV7 and HP1α interactions with CHAMP1. Thus, the exclusivity in forming CHAMP1/POGZ/HP1α or CHAMP1/POGZ/REV7 complexes is context-dependent. Taken together, our data suggest that the CHAMP1 complex may function in a REV7-independent pathway with heterochromatin. CHAMP1 complex promotes heterochromatin clustering at multiple genomic regions To investigate the potential function of the CHAMP1 complex in the formation of heterochromatin clusters, we initially assessed the localization of CHAMP1 and POGZ within heterochromatin foci by immunofluorescence. Remarkably, we observed a strong colocalization of CHAMP1 and POGZ with H3K9me3, which we used as a marker of heterochromatin, in human U2OS and mouse NIH-3T3 cells. Colocalization was further enhanced by ionizing radiation (IR) induced DSBs (Fig. [98]2a, b and Supplementary Fig. [99]2a, b). Consistently, the association between CHAMP1 and HP1α was stronger after cellular exposure to IR (Supplementary Fig. [100]2c). This interaction was not affected by the loss of REV7 interaction (i.e., the CHAMP1-2A mutant), supporting the idea that the CHAMP1 heterochromatin complex functions independently of REV7. Interestingly, radiation also induced the interaction of H3K9me3 with HP1α, but not with HP1β or HP1γ (Supplementary Fig. [101]2d), suggesting a specific role of HP1α in the heterochromatin enrichment at DSBs. Fig. 2. The CHAMP1 complex is required for the nuclear distribution of H3K9me3. [102]Fig. 2 [103]Open in a new tab a Representative immunofluorescence (IF) confocal images of the colocalized CHAMP1 and H3K9me3 foci with/without IR (5 Gy) treatment in U2OS cells. DAPI was used to stain the nuclei. Scale bar, 5 µm. Three independent experiments were performed with similar results. b Square analysis of the intensity and colocalization of H3K9me3 and CHAMP1 foci in (a). c Representative IF images of H3K9me3 foci in WT, sgCHAMP1, and sgPOGZ U2OS cells. DAPI was used to stain the nuclei. Scale bar, 5 µm. d Quantification of H3K9me3 foci in (c). Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. e Representative IF images of HP1α foci in WT, sgCHAMP1 and sgPOGZ U2OS cells. DAPI was used to stain the nuclei. Scale bar, 5 µm. f Quantification of HP1α foci in (e). Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. g Representative IF-FISH confocal images of H3K9me3 colocalization with centromere (Cent) in WT, sgCHAMP1, and sgPOGZ U2OS cells. DAPI was used to stain the nuclei. Scale bar, 5 µm. h Quantification of H3K9me3-Centromere colocalizations in (g). Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. i Representative IF confocal images of H3K9me3-TRF2 colocalization in WT, sgCHAMP1, and sgPOGZ U2OS cells. TRF2 was used as a telomere maker. DAPI was used to stain the nuclei. Scale bar, 5 µm. j Quantification of H3K9me3 colocalization with TRF2 in (i). Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. k ChIP-qPCR analyses showing relative enrichment levels of H3 (Top) and H3K9me3 (Bottom) at telomeres. Level of H3K9me3 was normalized to total H3 levels (H3K9me3/H3). Data are presented as mean values ±SD, n = 3 independent experiments. P values were calculated using a two-tailed Student’s T test. Source data are provided as a Source Data file. Notably, the CRISPR-knockout (KO) of CHAMP1 or POGZ in U2OS cells resulted in a significant reduction in the formation of H3K9me3 and HP1α foci (Fig. [104]2c–f). However, the KO of CHAMP1 or POGZ did not affect the overall expression level or chromatin binding of H3K9me3 and HP1α (Supplementary Fig. [105]2e, f, lane 9–12). To further determine whether the CHAMP1 heterochromatin complex promotes H3K9me3 accumulation at centromeres and telomeres, well-known heterochromatin sites, we next performed immunofluorescence with a centromere marker (Cent FISH) or a telomere marker (anti-TRF2 antibody)^[106]47. As expected, we observed a significant loss of H3K9me3 foci at both centromeres and telomeres, but not H3K9me2 foci, following the deletion of the CHAMP1 or POGZ in U2OS cells (Fig. [107]2g–j, Supplementary Fig. [108]2g, h). Consistently, chromatin immunoprecipitation (ChIP)-qPCR also demonstrated a decreased level of H3K9me3 at telomeres in CHAMP1 or POGZ KO U2OS cells (Fig. [109]2k). Taken together, the CHAMP1 complex plays a role in assembling the H3K9me3 heterochromatin mark at multiple sites in the genome, including sites with damage-induced DSBs as well as at centromeres and telomeres. Recent studies on telomeres in ALT tumor cells suggest a distinctive heterochromatin structure characterized by high levels of H3K9me3, low levels of ATRX, and elevated HDR processes at telomeres^[110]48–[111]50. We, therefore, decided to investigate the possible role of the CHAMP1 complex in the maintenance of these ALT telomere structures and HDR repair. We first examined the possible localization of the CHAMP1 complex at telomeres in ALT U2OS cells. Notably, we observed a strong colocalization of CHAMP1 and POGZ with telomere foci, particularly with large telomere foci, in U2OS cells (Supplementary Fig. [112]3a, b). This suggests that the CHAMP1 complex may play a role in promoting the formation of telomere clusters. DSBs at telomeres are known to initiate long-range movements and clustering of telomeres, and these processes are critical for homology-directed telomere synthesis^[113]51. Accordingly, we evaluated the response of U2OS cells to engineered DSBs at telomeres, using the TRF1-FokI^[114]51,[115]52. TRF1-FokI induced telomeric DSBs (Supplementary Fig. [116]3c, d), resulting in telomere clustering (Fig. [117]3a, b), as measured by an increase in the average size of telomere foci in ALT cells and a decrease in the total number^[118]51 (Fig. [119]3c, d). Interestingly, this DSB-induced telomere clustering was significantly diminished in CHAMP1 or POGZ KO U2OS cells (Fig. [120]3c, d). Depletion of CHAMP1 or POGZ also led to a decrease in the hallmarks of ALT recombination, including a reduction in ALT-associated PML bodies (APBs)^[121]14,[122]53 (Fig. [123]3e, f), which are sites of telomeric clustering and recombination^[124]15,[125]54, and a decrease of non-S phase telomeric DNA synthesis^[126]16 (Fig. [127]3g–i). Taken together, our results demonstrate that the CHAMP1 complex plays a critical role in facilitating damage-induced telomere clustering and in maintaining ALT activity (Fig. [128]3j). Fig. 3. The CHAMP1 Complex promotes break-induced telomere clustering and synthesis. [129]Fig. 3 [130]Open in a new tab a Schematic of U2OS-TRF1-FokI cell and time course of IF/FISH following Dox/4-OHT treatment. b Representative FISH confocal images with telomere probe (green) treated with/without Dox-induced TRF1-FokI in U2OS cells. DAPI was used to stain the nuclei. Scale bar, 5 µm. c, d CHAMP1 and POGZ KO U2OS cells were examined by telomere FISH after TRF1-FokI treatment (a). Average telomere (Telo) foci size (c) and telomere number (d) per nucleus were calculated using ImageJ. Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. e PML immunostaining combined with telomere FISH in wild-type (WT), sgCHAMP1, and sgPOGZ U2OS cells after Dox-induced TRF1-FokI treatment (a). Colocalization of PML foci and telomere signals (APBs) are shown as indicated. DAPI was used to stain the nuclei. Scale bar, 5 µm. f Quantification of APBs shown in (e). Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. g Schematic of U2OS-TRF1-FokI cell and time course of IF assay following Dox/4-OHT and EdU treatment. h Representative confocal images of immunostaining of TRF2 and EdU foci in S phase and non-S phase of U2OS cells following the treatment in (g). DAPI was used to stain the nuclei. Scale bar, 5 µm. i Quantification of the colocalized TRF2 and EdU foci in WT, sgCHAMP1, and sgPOGZ U2OS cells. Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. j A cartoon showing that CHAMP1 complex promotes telomere clustering, APBs formation and facilitates Break-induced new telomeric DNA synthesis. Source data are provided as a Source Data file. CHAMP1 complex promotes HDR at ALT telomeres We next used the PICh (Proteomics of Isolated Chromatin Segments) method to identify the recruitment of telomere-associated proteins following DSBs, in the presence or absence of the CHAMP1 heterochromatin complex (Fig. [131]4a and Supplementary Fig. [132]4a). Consistent with previous observations^[133]52, multiple DNA repair- and telomere maintenance-related biological processes were significantly upregulated at telomeres following TRF1-FokI induction (Supplementary Fig. [134]4b). CRISPR KO of CHAMP1 resulted in a decrease of other CHAMP1 complex proteins (i.e., POGZ and HP1α) at telomeres with DSBs, as well as a decrease of many proteins associated with HR (e.g., RAD51, BRCA1, and BARD1), as measured by mass spectrometry (Fig. [135]4b). These results are consistent with the known role of CHAMP1 and POGZ in HR repair^[136]34–[137]36. Additionally, CHAMP1 depletion also resulted in a reduction in proteins involved in BIR or break-induced telomere synthesis (BITS)^[138]52,[139]55 (i.e., POLD3, POLD4, RFC, and PCNA) (Fig. [140]4b). Western blot analysis of PICh-purified telomeres revealed an enrichment of the CHAMP1 complex (i.e., CHAMP1 and HP1α) and BITS proteins (i.e., POLD3, RAD18 and PCNA) at telomeres in U2OS cells following TRF1-FokI induction, an effect that was reduced after depletion of CHAMP1 (Fig. [141]4c, lanes 9–12). Immunofluorescence further confirmed the reduction of CHAMP1 complex (i.e., POGZ and HP1α) and H3K9me3 at telomeres in CHAMP1 KO U2OS cells with or without DSB induction (Fig. [142]4d–g and Supplementary Fig. [143]4c, d), indicating that the lower number of foci upon DSB induction may stem from a reduced baseline number. Consistently, POGZ depletion also resulted in a reduction of proteins associated with HR and BITS at telomeres with DSBs (Supplementary Fig. [144]4e). Consistent with the local reduction in HR proteins, KO of CHAMP1 or POGZ in U2OS cells resulted in a reduction in HR activity at telomeres, as measured by RAD51 and pRPA2(S33) foci at damaged telomeres (Fig. [145]4h, i and Supplementary Fig. [146]4f, g). Taken together, these results support a model in which the CHAMP1 complex promotes HDR, potentially through the clustering of heterochromatin at telomeres. Fig. 4. The CHAMP1 complex establishes telomeric heterochromatin and promotes the recruitment of HR and BIR factors. [147]Fig. 4 [148]Open in a new tab a Schematic of the PICh approach used to define the telomere-specific proteome in WT and sgCHAMP1 U2OS cells after TRF1-FokI (±Dox/4-OHT) treatment. Fractionated and precleared chromatin was hybridized to a biotinylated telomeric probe and subsequently captured on magnetic beads. Telomere-associated proteins were analyzed by silver staining, western blot analysis and mass spectrometry. PFA, paraformaldehyde. b Scatterplot of the telomere-specific DSB response proteome profile enriched by PICh in wild-type (WT) and sgCHAMP1 U2OS cells induced with TRF1-FokI. Mass Spectrometry data was shown in Supplementary Data [149]1. Scatterplot of log2 FC[(+Dox+1)/(−Dox+1) of total peptide number] from two independent experiments. FC, fold change. Homologous recombination (HR) pathway proteins with Log2 FC > 1 in WT were highlighted in red. Break-induced replication (BIR) pathway proteins with Log2 FC > 1 in WT were highlighted in green. CBX5 (HP1α) was highlighted in orange. Telomere-specific binding proteins (Shelterin) and POGZ were highlighted in blue. The red dotted line indicates the same value of log2 FC(+Dox/−Dox) in both WT and sgCHAMP1 cells. c Western blot from PICh experiments in U2OS with/without Dox-induced TRF1-FokI. Telomere-associated proteins are blotted with the indicated antibodies. d Representative IF confocal images depicting the colocalization of HP1α (Green) with TRF1-FokI (T1-F1, mCherry tagged, Red) in WT and sgCHAMP1 U2OS cells induced with/without TRF1-FokI for 2 hours. DAPI (blue) was used to stain the nuclei. Scale bar, 5 µm. e Quantification of HP1α-TRF1-FokI colocalization events in (d). Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. f Representative IF confocal images depicting the colocalization of H3K9me3 (Green) with TRF1-FokI (T1-F1, mCherry tagged, Red) in WT and sgCHAMP1 U2OS cells induced with/without TRF1-FokI for 2 hours. DAPI (blue) was used to stain the nuclei. Scale bar, 5 µm. g Quantification of H3K9me3-TRF1-FokI colocalization events in (f). Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. h Representative IF confocal images depicting the colocalization of RAD51 (Green) with TRF1-FokI (T1-F1, mCherry tagged, Red) in WT, sgCHAMP1 and sgPOGZ U2OS cells induced with/without TRF1-FokI for 2 hours. DAPI (blue) was used to stain the nuclei. Scale bar, 5 µm. i Quantification of RAD51-TRF1-FokI colocalization events in (h). Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. Source data are provided as a Source Data file. CHAMP1 complex binds and recruits SETDB1 to telomeric heterochromatin clusters To further evaluate the regulation of telomere heterochromatin deposition, we examined H3K9 methyltransferase levels and activity. Interestingly, SETDB1, but not other H3K9 methyltransferases, was recruited to telomeres after TRF1-FokI induction in U2OS cells, whereas this recruitment was not observed in CHAMP1- and POGZ-depleted U2OS cells (Fig. [150]5a). Immunofluorescence confirmed the reduction in SETDB1 at telomeres following TRF1-FokI induction in CHAMP1 or POGZ KO U2OS cells (Fig. [151]5b, c). Based on previously published data, indicating a POGZ and SETDB1 interaction^[152]56, we reasoned that the CHAMP1 complex might directly recruit SETDB1 to telomeres. A strong interaction between SETDB1 and the Zn-fingers (type 2–7) of POGZ was predicted (iPTM = 0.85) (Fig. [153]5d). Furthermore, an N-terminal deletion of CHAMP1, causing loss of the interaction with POGZ, also exhibited reduced interaction with SETDB1 (Fig. [154]5e). These results suggest that the CHAMP1 complex may recruit SETDB1 to telomeres through POGZ (Supplementary Fig. [155]5a). Fig. 5. The CHAMP1 complex promotes the recruitment of SETDB1 to telomeres. [156]Fig. 5 [157]Open in a new tab a Scatterplot of Log2 (+Dox+1)/(−Dox+1) of total peptide number from the telomere-specific DSB response proteome profile enriched by PICh in wild-type (WT), sgCHAMP1 and sgPOGZ U2OS cells induced with TRF1-FokI. Mass Spectrometry data was shown in Supplementary Data [158]2. SETDB1 is highlighted. b Representative IF confocal images of SETDB1 (Green) colocalization with telomeres with double-strand breaks (T1-F1, mCherry, Red) in wild-type (WT), sgCHAMP1 and sgPOGZ U2OS cells induced with TRF1-FokI for 2 h. The arrow indicates SETDB1 and TRF1-FokI colocalization. DAPI was used to stain the nuclei. Scale bar, 5 µm. c Quantification of (a) for a number of SETDB1-mCherry colocalization events. Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. d (Top) AlphaFold2-Multimer (AF2)-predicted structural model of POGZ_ZnFcore (peach)-SETDB1_Cter (cyan) complex. (Bottom) The AF2-predicted model of POGZ_ZnFcore-SETDB1_Cter complex is colored based on the confidence in the model prediction (100-high (blue) and 50-low (red)) and corresponding PAE matrix showing the confidence in the predicted interaction between POGZ_ZnFcore and SETDB1_Cter domains along with iPTM scores for the AF2 prediction. Unstructured loops connecting structured domains in the model have been removed for clarity. e Western blot showing GFP-immunoprecipitation of GFP-vector control, GFP-CHAMP1 WT or deletion mutants, and the co-immunoprecipitation of endogenous SETDB1 and POGZ. GAPDH acts as a negative control for IP. Immunoblots are representative of two independent experiments. f Representative IF-FISH confocal images of CHAMP1 colocalization with telomeres in wild-type (WT) and sgSETDB1 U2OS cells. Arrow indicates CHAMP1-telomere colocalization. DAPI (blue) was used to stain the nuclei. Scale bar, 5 µm. g Quantification of (f) for number of CHAMP1-telomere colocalization events. Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. h SETDB1 KO TRF1-FokI U2OS cells were examined by FISH using a FITC-labeled telomere probe after DOX/4-OHT treatment. Average telomere foci size per nucleus was calculated using ImageJ. Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. i PML immunostaining combined with telomere FISH in SETDB1 KO U2OS cells after DOX/4-OHT treatment. Colocalization of PML foci and telomere signals (APBs) are shown. Data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. j A cartoon showing that the recruitment of SETDB1 by CHAMP1 complex could potentially facilitates a positive feedback loop that boosts the formation of heterochromatin clusters. Source data are provided as a Source Data file. Recently, it has been shown that ALT activity relies on a high level of H3K9 trimethylation, facilitated by SETDB1, and on the recruitment of recombination factors^[159]49. Consistently, KO of SETDB1 resulted in a decrease of H3K9me3 foci at telomeres (Supplementary Fig. [160]5b, c). Importantly, SETDB1 depletion decreased CHAMP1 recruitment to telomeres (Fig. [161]5f, g). Furthermore, SETDB1 depletion in U2OS cells led to reduced damage-induced telomere clustering and APBs (Fig. [162]5h, i). Collectively, these findings suggest that the CHAMP1 heterochromatin complex recruits SETDB1, thereby promoting a positive feedback loop that enhances the formation of telomere heterochromatin clusters (Fig. [163]5j). CHAMP1 and HP1α are epistatic in the regulation of heterochromatin clustering Since HP1α is a critical subunit of the CHAMP1 heterochromatin complex, we next examined its role in the regulation of telomeric HR and clustering. Consistent with previous findings, HP1α is required for APBs formation^[164]57. HP1α knockdown resulted in a significant decrease of damage-induced telomere clustering and APBs formation (Fig. [165]6a, b and Supplementary Fig. [166]6a). HP1α knockdown did not result in a further decrease in damage-induced telomere clustering and APBs in CHAMP1 KO U2OS cells (Fig. [167]6a, b and Supplementary Fig. [168]6a), indicating that CHAMP1 and HP1α are epistatic for APB formation and telomere clustering. Consistent with the requirement of HP1α dimerization in the function of the CHAMP1 complex, the wild-type HP1α protein, but not the HP1α-I165E dimerization mutant, restored telomere clustering and APBs in cells with an HP1α knockdown (Fig. [169]6c, d and Supplementary Fig. [170]6b). Furthermore, the C-terminal deletion mutant of CHAMP1, which lacks HP1α binding, and the N-terminus deletion mutant of CHAMP1, which lacks POGZ binding, did not restore RAD51 foci at telomeres or telomere clustering after CHAMP1 depletion (Fig. [171]6e, f and Supplementary Fig. [172]6c). These findings highlight the critical importance of the multi-subunit CHAMP1 heterochromatin complex in telomeric HR and clustering. Fig. 6. HP1α and CHAMP1 are epistatic in the regulation of telomeric HR and clustering. [173]Fig. 6 [174]Open in a new tab a, b Wild-type (WT) and sgCHAMP1 U2OS-TRF1-FokI cells were treated with siControl or siHP1α, followed by examination of FISH using a telomere probe, or PML combined with telomere IF-FISH staining after DOX/4-OHT treatment. a Average telomere foci size per nucleus was calculated using ImageJ. b Quantification of colocalization of PML foci and telomere signals (APBs). c, d WT, sgHP1α, and sgHP1α U2OS cells with ectopically expressed HP1α WT or mutant I165E were examined with FISH using a telomere probe, or PML combined with telomere IF-FISH staining using a FITC-labeled telomere probe after DOX/4-OHT treatment. c Average telomere foci size per nucleus was calculated using ImageJ. d Quantification of colocalization of PML foci and telomere signals (APBs). e, f WT, sgCHAMP1, and sgCHAMP1 U2OS cells with ectopically expressed CHAMP1 WT or mutants (ΔN, 2 A and ΔC) were examined with IF using anti-RAD51 or FISH using a telomere probe. e Quantification of RAD51-TRF1-FokI colocalization events. f Average telomere foci size per nucleus was calculated using ImageJ. For all of the foci analysis, data are presented as mean values ±SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using a two-tailed Student’s t test. Source data are provided as a Source Data file. The CHAMP1 heterochromatin complex has a specific role in telomeric HR repair The interaction of CHAMP1 and REV7 appears to be dispensable for the telomere HR and clustering functions (Fig. [175]6e, f and Supplementary Fig. [176]6c). Interestingly, immunoprecipitation of REV7 pulls down CHAMP1-WT and POGZ, but does not coimmunoprecipitate HP1α or other heterochromatin proteins. (Fig. [177]1e, g, h). Moreover, the CHAMP1-2A mutant protein, which has a disrupted binding site for REV7, can still assemble into the CHAMP1 complex and at least partially promote heterochromatin cluster formation and promote HR activity (ie, RAD51 foci at sites of DSBs) (Fig. [178]6e, f and Supplementary Fig. [179]6c). Importantly, the knockdown of REV7 in both CHAMP1-WT and -depleted cells results in a slight increase in HR at telomeric heterochromatic regions (Supplementary Fig. [180]6d,e). However, the KO of CHAMP1 in ERV7-depleted cells still significantly decreases HR activity (Supplementary Fig. [181]6d, e), indicating that a REV7-independent CHAMP1 complex still plays an important role in HR repair in heterochromatin. Furthermore, we observed a reduced level of POGZ in CHAMP1 KO cells. To determine whether the effects seen upon CHAMP1 depletion were caused by the reduced level of POGZ, we examined the POGZ chromatin binding with multiple CHAMP1 mutants. Interestingly, although all CHAMP1 mutants exhibited HR defects, only the N-terminal mutants (ΔN, ΔNΔC, and QLT/RRR) reduced POGZ chromatin binding. The 2 A and delta C mutants of CHAMP1 did not exhibit reduced POGZ chromatin binding (Supplementary Fig. [182]6f). Taken together, these results suggest that the effects of CHAMP1 depletion are not due to the loss of POGZ. Instead, the entire CHAMP1 complex plays a role in heterochromatin HR repair. These results demonstrate that the activity of the REV7-independent CHAMP1 heterochromatin complex is essential for heterochromatic clustering and HR repair in heterochromatin. Patients with CHAMP1 syndrome display impaired heterochromatin clustering and defects in HR repair We next examined EBV-immortalized lymphoblast cell lines derived from patients with CHAMP1 syndrome, resulting from an inherited mutation in one allele of the CHAMP1 gene. The cells were obtained from the Coriell mutant cell line repository (Fig. [183]7). These cell lines express mutant (truncated) CHAMP1 proteins (Fig. [184]7a, and Supplementary Fig. [185]7a), as predicted by the genomic sequencing of the CHAMP1 gene. These truncated proteins, lacking the C-terminal HP1α binding domain, may act as dominant negative proteins or lead to protein instability and haploinsufficiency. Interestingly, these CHAMP1 patient-derived cell lines exhibit a defect in heterochromatin clustering, as indicated by a decrease of H3K9me3 foci (Fig. [186]7b, c) compared to wild-type control cells. The CHAMP1 patient-derived cell lines also exhibit a defect in RAD51 foci assembly following IR (Fig. [187]7d, e), consistent with a deficiency in HR, and exhibit heightened sensitivity to both IR and PARP inhibitors (Fig. [188]7f, g). However, significant differences in the cell cycle of these cell lines, compared to wild-type cells (Supplementary Fig. [189]7b), were not observed. Taken together, peripheral blood lymphocytes from individuals carrying CHAMP1 mutations exhibit defects in heterochromatin clustering and deficiencies in HR, providing a possible functional diagnostic assay for the CHAMP1 syndrome. Fig. 7. Peripheral blood lymphocytes from patients with CHAMP1 syndrome display impaired heterochromatin clustering and defects in HR repair. [190]Fig. 7 [191]Open in a new tab a Schematic of CHAMP1 mutants from patients with CHAMP1 syndrome. b EBV-immortalized lymphoblast cell lines derived from several CHAMP1 patients (Coriell) were examined by immunostaining with H3K9me3 antibody. Representative images show the H3K9me3 foci. DAPI was used to stain the nuclei. Scale bar, 5 µm. c Quantification of H3K9me3 foci of the CHAMP1 patient lymphoblast cells in (b). Error bars indicate SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using one-way ANOVA. d Representative images of RAD51 foci formation in normal and CHAMP1 patient lymphoblast cells following with/without 5 Gy IR treatment. DAPI was used to stain the nuclei. Scale bar, 5 µm. e Quantification of RAD51 foci of the CHAMP1 patient lymphoblast cells following 5 Gy IR treatment. Error bars indicate SD, n = 3 independent experiments. More than 50 cells were counted for each experiment. P values were calculated using one-way ANOVA. f 3-day cytotoxicity analysis of the normal and CHAMP1 patient lymphoblast cells treated with various doses of radiation. Cell viability was detected by CellTiter Glo (Promega). Error bars indicate SD, n = 3 independent experiments. g 3-day cytotoxicity analysis of the normal and CHAMP1 patient lymphoblast cells treated with various doses of Olaparib. Cell viability was detected by CellTiter Glo (Promega). Data are presented as mean values ±SD, n = 3 independent experiments. Source data are provided as a Source Data file. Discussion Taken together, our studies show that the CHAMP1 heterochromatin complex, consisting of CHAMP1, POGZ, and HP1α, has a direct role in heterochromatin assembly and Homology-Directed DNA Repair. The complex binds to H3K9me3 in heterochromatin sites, including centromeres and telomeres, and contributes to the H3K9 trimethylation at these sites. The complex is also recruited to heterochromatic sites with DSBs generated by IR. Moreover, the CHAMP1 heterochromatin complex plays a role in HDR at these heterochromatic sites. Our results support the presence of two distinct CHAMP1/POGZ complexes. One is REV7-dependent, and the other is REV7-independent. We previously showed that the CHAMP1/POGZ complex promotes HR at DNA DSBs by binding to REV7 and limiting the binding of REV7 to proteins in the Shieldin complex (i.e., the REV7, SHLD1, SLD2, SHLD3 complex). The Shieldin complex is known to block the resection of DSBs and thereby limit the level of HR repair. While this HR mechanism primarily occurs in euchromatic regions, it may also occur to some extent in heterochromatic regions. However, in the current study, we identify another CHAMP1 complex, the CHAMP1 heterochromatin complex, whose function was previously unknown. This latter complex is enriched in heterochromatin and excludes the CHAMP1 binding partner REV7. Here, we show that this complex interacts directly with HP1α and promotes the H3K9me3 marks in heterochromatin. Moreover, this complex includes the SETDB1 methyltransferase. The complex is required for the recruitment of HDR proteins to heterochromatic regions of the chromosome. Loss of the CHAMP1 heterochromatin complex results in a decrease in the repair of TRF1-FokI mediated DSBs at telomeres. Knockdown of CHAMP1 complex subunits, or mutation of individual subunits preventing complex assembly, reduced pRPA and RAD51 foci at these heterochromatic sites. Evaluation of the CHAMP1-2A mutant protein, which fails to bind to REV7^[192]36 is especially informative. CHAMP1-2A protein does not enable HDR repair by preventing the binding of REV7 with the Shieldin complex^[193]36. However, the CHAMP1-2A mutant protein is still functional for HDR repair in heterochromatin through its interaction in the REV7-independent CHAMP1 heterochromatin complex. As a subunit of the CHAMP1 heterochromatin complex, the CHAMP1-2A mutant protein still allows heterochromatin clustering, recruitment of HDR proteins to heterochromatin, and repair of DSBs in heterochromatin (i.e., formation of RAD51 foci). Our study also demonstrates that the CHAMP1 heterochromatin complex is especially important to heterochromatin clustering and HDR repair at specific heterochromatic sites. For instance, the telomeres of ALT tumor cells are known to have high levels of H3K9me3 and high HDR activity, as well as high levels of replication stress. ALT telomeres appear to have a unique heterochromatin environment, lacking the protein ATRX^[194]49,[195]58. Since SETDB1 establishes the H3K9me3 mark on telomeres^[196]49, it would also be interesting to determine whether the localization of CHAMP1 and POGZ on telomeres can also be observed in mouse embryonic stem (mES) cells or primarily in ALT cells. The CHAMP1 heterochromatin complex may be required for the assembly and maintenance of ALT telomeres. The knockdown or KO of the CHAMP1 complex results in decreased DSB-induced telomere cluster formation and reduced HDR activity, coincident with a loss of telomere-associated histone H3 lysine 9 trimethylation (H3K9me3). Further studies will be required to determine the role of the CHAMP1 complex in the maintenance of telomere length. The CHAMP1 heterochromatin complex may play additional roles in heterochromatin structures. For instance, the complex may be regulated by ATM kinase^[197]2 or may function in HDR repair at the periphery of the cell nucleus^[198]9,[199]18. The complex may play a role in specific signaling events required for assembling heterochromatin at stressed replication forks^[200]59,[201]60. Future studies will be also required to assess the role of the CHAMP1 heterochromatin complex at these other heterochromatic sites, such as centromeres and DNA replication forks. A reduction in H3K9me3 is known to result from a mislocalization or inactivation of the SETDB1, a methyltransferase known to trimethylate H3K9^[202]61. Consistent with these findings, our AF2 analysis, immunoprecipitation studies, and PICh results revealed an interaction of SETDB1 with the POGZ subunit of the complex, required for the recruitment of SETDB1 to sites of heterochromatin. Taken together, the CHAMP1 complex appears to be a critical determinant of heterochromatin clustering. Genetic ablation of a subunit of the complex, or a mutation in a subunit that disrupts complex assembly, may result in a failure to recruit SETDB1 and in loss of heterochromatin clustering. Interestingly, GFP-CHAMP1-deltaC prevents CHAMP1-HP1 binding while it does not affect the interaction between POGZ/CHAMP1 and SETDB1 (Fig. [203]5e). This suggests the association between POGZ/CHAMP1 and SETDB1 occurs independently of HP1 and chromatin association. However, given that ALT telomeres are longer and exhibit more replication stress-induced damage, it is possible that SETDB1 may be recruited to ALT telomeres through the CHAMP1-POGZ interaction. This could lead to an enhanced role of SETDB1 in maintaining heterochromatin integrity at these sites. Previous studies have suggested that tightly packed heterochromatin regions have an overall decrease in HDR activity^[204]3. Paradoxically, in our studies and in other recent reports^[205]49,[206]62,[207]63, the increase in H3K9 trimethylation or other modifications in the heterochromatin composition at telomeres may enhance HDR activity. Consistent with these findings, the CHAMP1 complex may promote HDR by functioning as a bridge between two heterochromatin nucleosomes, either intra- or interchromosomally, through its interaction with H3K9me3 in heterochromatin. This bridge may facilitate the formation of heterochromatin clusters, thereby creating an environment that promotes the recruitment and stabilization of HDR proteins. Finally, individuals with inherited heterozygous mutations in CHAMP1 present with a neurodevelopmental disorder characterized by intellectual disability, behavioral symptoms, and distinct dysmorphic features. Interestingly, individuals with de novo heterozygous mutations in the POGZ gene also exhibit a highly related neurodevelopment syndrome, known as the WSS^[208]25,[209]27,[210]28,[211]30,[212]31,[213]64. The mutant proteins encoded by pathogenic alleles of CHAMP1 have C-terminus truncations and perhaps dominant negative activity. These mutant proteins cannot assemble into functional CHAMP1 complexes. In fact, the CHAMP1 deltaC mutant, which mimics the CHAMP1 mutation found in GM27648 cells, retains its ability to bind REV7. These findings support the idea that CHAMP1 mutations impair heterochromatin formation without affecting REV7 binding. However, although the heterozygous CHAMP1 patient cells express a significant amount of CHAMP1-WT, their expression levels are considerably lower compared to normal cells. This could also indicate a dosage effect. Accordingly, lymphoblast cultures derived from individuals with CHAMP1 syndrome and expressing these mutant proteins have defects in heterochromatin formation and HR. These functional defects may also underlie the neurodevelopmental defects in these individuals. Lymphoblasts from these individuals have phenotypic characteristics similar to engineered human cells lacking the CHAMP1 complex. Taken together, these cellular characteristics could serve as a functional diagnostic screening tool for identifying individuals with CHAMP1 syndrome among the much larger number of individuals with autism spectrum disorders. In conclusion, we provide evidence that the CHAMP1 heterochromatin complex is critical for heterochromatin HDR repair, not only in tumor cells but also in normal tissue. Methods Cell lines U2OS and U2OS TRF1-FOKI cell lines were grown in DMEM/F12 (Thermo Fisher) with 10% FBS (sigma) and 1% penicillin-streptomycin (Thermo Fisher). 293 T, NIH-3T3, and HeLa S3 cell lines were grown in DMEM (Thermo Fisher) with 10% FBS (sigma) and 1% penicillin-streptomycin (Thermo Fisher). CHAMP1 patient lymphocyte cell lines from Coriell Institute were grown in RPMI1640 with 10% FBS (sigma) and 1% penicillin-streptomycin (Thermo Fisher). Cell lines were maintained in an incubator at 37 °C and 5% CO2 according to standard protocols. Cell lines were validated to be negative for mycoplasma contamination using the MycoAlert Plus Mycoplasma Detection Kit (Lonza) and the rapid MycoBlue Mycoplasma Detector (Vazyme). RNA interference, gene over-expression, and CRISPR-mediated gene deletions DNA transfections were performed using Lipofectamine LTX (Invitrogen), while siRNA knockdowns were executed using RNAiMax (Invitrogen), following the manufacturer’s guidelines. For single gene KO, sgRNAs targeting candidates indicate genes were either cloned into the pSpCas9 BB-2A-GFP (PX458) vector (GenScript) or introduced into the cell together with the Cas9 protein via electroporation (Lonza) according to the manufacturer’s protocol. After Cas9-gRNA PX458 plasmid transfection, GFP-positive cells were sorted using a BD FACSAria II cell sorter 48 hours post-transfection. GFP positive pool or single cells were screened for KOs by western blotting. Cellular fractionation and immunoblot analysis The cells were lysed using RIPA buffer supplemented with a cocktail of phosphatase and protease inhibitors from Roche. Cell lysates were separated by electrophoresis using NuPAGE 4–12% Bis-Tris gels (Invitrogen) and transferred onto nitrocellulose membranes. These membranes were then blocked with 5% BSA in TBST (Tris-buffered saline with Tween) and subsequently incubated with primary and secondary antibodies. The detection was accomplished using either chemiluminescence or fluorescence (LI-COR Biosciences). For chromatin extraction, chromatin-bound extracts were obtained using a subcellular protein fractionation kit from Thermo. Band intensities were quantified using ImageJ. Immunofluorescence and IF-FISH assays Cells were seeded onto glass coverslips placed in 24-well plates. Subsequently, they were either left untreated or exposed to 5 Gy of IR. Following 6 hours, the cells were collected by pre-extraction with 0.5% Triton X-100 for 5 minutes, followed by fixation with 4% paraformaldehyde for 10 minutes at 4 °C. After three PBS washes, a blocking step was carried out using 3% BSA in PBS for 1 hour at room temperature. This was followed by consecutive incubations with primary and secondary antibodies, conducted overnight at 4 °C and 1 hour at room temperature, respectively. For the IF-FISH assay, coverslips were first stained with the primary and secondary antibodies, fixed for 10 min at room temperature, and dehydrated in ethanol series. After denaturation at 85 °C for 5 min, coverslips were incubated with TelG-Cy3 or TelC-Alexa488 PNA probe (PNAbio) overnight at 37°C, then washed. Finally, the coverslips were mounted with DAPI (Vector Laboratories) and imaged using a Zeiss AX10 fluorescence microscope and Zen software. Telomere foci size and clustering were measured by ImageJ using a consistent threshold to images followed by binarization as described previously^[214]51. The sizes of the foci were quantified in square pixels for each telomeric focus within a nucleus, and the average size was computed for each analyzed nucleus. CellTiter-Glo assay To evaluate the sensitivity of CHAMP1 patient lymphocyte cells to PARP inhibitors (PARPi) and IR, the short-term CellTiter-Glo survival assays were performed as described previously^[215]65. For IR sensitivity, cells were exposed to an indicated dose of IR and then plated in 96-well plates at a density of 800 cells per well. For PARPi sensitivity, cells were initially seeded in 96-well plates at a density of 800 cells per well and then treated with drugs at the indicated concentrations after 12 hours. Three days later, cellular viability was assessed using CellTiter-Glo (Promega). Survival at each dose of IR or drug concentration was calculated as a percentage relative to the corresponding untreated control. Immunoprecipitation After transfection for 48 hours, 293 T cells were subsequently collected and subjected to lysis using NETN lysis buffer containing a proteinase and phosphatase inhibitor cocktail (Thermo, 1:100) for 30 minutes on ice. Following this, the lysed samples were incubated overnight at 4 °C with an antibody-bead conjugate, which consisted of GFP-Trap_A (Chromotek). The beads were then thoroughly washed four times with NETN buffer, and the immunoprecipitated materials were eluted by boiling. Western blot analysis was conducted to detect the immunoprecipitates, and the intensities of the resulting bands were quantified using ImageJ. ChIP and qPCR analysis ChIP assays were performed, as previously described. Initially, cells were treated with 1% formaldehyde for 10 minutes at room temperature to cross-link proteins to DNA, and the reaction was then halted by adding glycine to a final concentration of 0.125 M for 5 minutes. For ChIP assay, the SimpleChIP Kits (Cell Signaling Technology # 9003S) were used. 5 μg H3 and H3K9me3 antibodies were used for each ChIP reaction. Purified ChIP DNA was used as a template for qPCR using the primers corresponding to telomeric repeats. qPCR experiments were performed using the QuantStudio™ 7 Flex Real-Time PCR System (ABI) with 96-well PCR plates and sealing films (Vazyme). Telomeric DNA synthesis in G2 phase To visualize telomeric DNA synthesis, U2OS cells were synchronized in G2 by treatment with 15 μM CDK1i (RO-3306) for 16 h with or without the addition of Doxycycline. Cells were incubated with 20 µM EdU for an additional 2 hours, with or without 4-OH Tamoxifen, and then fixed with 4% PFA for 10 min at RT. Cells were permeabilized with 0.2% Triton X-100 in PBS for 10 min at RT, blocked with 10% normal goat serum for 1 h at 4 °C then incubated overnight with anti-TRF2 antibody. Cells were then washed with 0.1% Triton X-100 in PBS, incubated with fluorescently-labeled secondary antibody and EdU was labeled with fluorescent dye using Click-iT EdU kit (Invitrogen) according to the manufacturer’s protocol. DNA stained with DAPI and z-stack images were acquired using a Zeiss AxioObserver microscope at ×63 magnification. Images were analyzed using ImageJ software and the number of EdU+ TRF2 foci was assessed in at least 150 cells from three independent experiments. Proteomics of isolated chromatin segments (PICh) Telomere-associated proteins were captured as previously described using PICh protocol^[216]52,[217]66,[218]67 with the following modifications. U2OS (~10^9 cells) TRF1-FokI was induced with Doxycycline (40 ng ml^−1, Sigma-Aldrich, D9891) for 18hrs, then 4-Hydroxytamoxifen (1 µM, Sigma-Aldrich, H7904) was added for another 2 hrs. Cells were fixed with 4% formaldehyde (Sigma-Aldrich) for 45 min with shaking at room temperature, washed three times with ice-cold 1× PBS, and scrapped down and collected in 1× PBS + 0.05% Tween 20. The pellets were washed and dounced in sucrose solution (0.3 M Sucrose, 10 mM HEPES-NaOH pH 7.9, 1% Triton-X-100, 2 mM Magnesium Acetate). Cell pellets were resuspended in Triton solution (1× PBS, 0.5% Triton X-100, 1 mM PMSF, and 0.25 mg ml^−1 RNase A (Qiagen, 19101)) overnight on a rotator at 4 °C, then washed with ice-cold 1× PBS for three times and with LB4 lysis buffer (50 mM Tris-HCl pH 8.0, 200 mM NaCl, 20 mM EDTA-NaOH pH 8.0, 1% SDS) for twice. Cell pellets resuspended in lysis buffer containing 1 mM PMSF and sonicated with BRANSON Digital Sonifier with the settings at Amplitude: 70%; Pulse cycle: 15 sec On, 45 sec Off; Total process time: 7.5 min. Samples were collected and heated at 58°C for 5 min and centrifuged at 16,000 × g for 10 min at room temperature. The soluble fraction was collected and precleared with 1 mL Streptavidin Agarose (EMD Millipore, 69203-3) for 3hrs at room temperature and then passed through Sephacryl S-400 HR column (GE Healthcare Life Sciences, 17060901) at 750 × g for 5 min. 0.01% of the sample was saved as “Input” for western blot. Soluble fractions were hybridized with 1500 pmol desthiobiotin labeled 2’-Fluor-RNA telomeric probes (t-Desthiobiotin TEG-Spacer18-Spacer18-UUAGGGUUAGGGUUAGGGUUAGGGUUAGGGUUAGGGUUAGGGt) in a thermocycler (25 °C for 3 min; 80 °C for 5 min; 37 °C for 60 min; 60 °C for 3 min; 37 °C for 30 min; 60 °C for 3 min; 37 °C for 30 min; then keep at 25 °C). Hybridized chromatin was collected and span at 16,000 × g for 15 min at room temperature and then incubated with 900 µL Dynabeads™ MyOne™ Streptavidin C1 (Thermo Fisher, 65002) overnight at room temperature. Then bound chromatin was immobilized on a magnetic stand. 0.01% of unbound fraction was taken as the “Unbound” for Western-blot, followed by washing five times with LB3JD buffer (10 mM HEPES-NaOH pH 7.9, 100 mM NaCl, 2 mM EDTA-NaOH pH 8.0, 1 mM EGTA-NaOH pH 8.0, 0.2% SDS, 0.1% Sodium Sarkosyl) at room temperature and one wash with LB3JDS buffer (10 mM HEPES-NaOH pH 7.9, 30 mM NaCl, 2 mM EDTA-NaOH pH 8.0, 1 mM EGTA-NaOH pH 8.0, 0.2% SDS, 0.1% Sodium Sarkosyl) at 42 °C at 1000 rpm for 5 min. The telomere and associated proteins on beads were eluted twice with 450 µL elution buffer (75% LB3JD buffer + 25% D-biotin (Invitrogen, [219]B20656)). The eluted samples, as well as the “Input” and “Unbound” fractions, were then precipitated with TCA precipitation protocol, decrosslinked in 250 mM Tris-HCl pH 8.8, 2% SDS, 0.1 M 2-Mercaptoethanol, and boiled with NuPAGE LDS Sample Buffer (4×, Invitrogen). The samples were then analyzed by silver staining, western blot, and mass spectrometry (Taplin Biological Mass Spectrometry Facility at Harvard Medical School). Mass spectrometry and data analysis Excised gel bands after PICh were cut into ~1 mm^3 pieces. Gel pieces were then subjected to a modified in-gel trypsin digestion procedure^[220]68. Gel pieces were washed and dehydrated with acetonitrile for 10 min. followed by removal of acetonitrile. Pieces were then completely dried in a speed vac. Rehydration of the gel pieces was with 50 mM ammonium bicarbonate solution containing 12.5 ng/µl modified sequencing-grade trypsin (Promega, Madison, WI) at 4 °C. After 45 min., the excess trypsin solution was removed and replaced with 50 mM ammonium bicarbonate solution to just cover the gel pieces. Samples were then placed in a 37 °C room overnight. Peptides were later extracted by removing the ammonium bicarbonate solution, followed by one wash with a solution containing 50% acetonitrile and 1% formic acid. The extracts were then dried in a speed-vac (~1 hr). The samples were then stored at 4 °C until analysis. On the day of analysis, the samples were reconstituted in 5–10 µl of HPLC solvent A (2.5% acetonitrile, 0.1% formic acid). A nano-scale reverse-phase HPLC capillary column was created by packing 2.6 µm C18 spherical silica beads into a fused silica capillary (100 µm inner diameter x ~ 30 cm length) with a flame-drawn tip^[221]69. After equilibrating the column each sample was loaded via a Famos auto sampler (LC Packings, San Francisco, CA) onto the column. A gradient was formed and peptides were eluted with increasing concentrations of solvent B (97.5% acetonitrile, 0.1% formic acid). LC-MS/MS analysis was performed using a Velos Orbitrap Pro ion-trap mass spectrometer (Thermo Fisher Scientific, Waltham, MA). Peptides were subjected to electrospray ionization and then entered the mass spectrometer, where they were detected, isolated, and fragmented to generate tandem mass spectra. Peptide sequences (and hence protein identity) were determined by matching the acquired fragmentation pattern against protein databases using the Sequest software (Sequest ver 28 rev 13, Thermo Fisher Scientific, Waltham, MA)^[222]70. The output data were processed using in-house software (GFY Core Version 3.12 - Module search Version 3.12). All databases included reversed versions of all sequences, and the data were filtered to achieve a peptide-level false discovery rate of 1% to 2%. The protein sequence database used was downloaded from Uniprot (human database downloaded on 6 January 2023, [223]https://www.uniprot.org/uniprotkb?query=proteome:UP000005640). For each protein after PICh, log[2] foldchange (LFC) of the total peptide count between +Dox and −Dox samples was computed separately for each replicate of WT and sgCHAMP1 USOS cells. The average LFC was then computed from two independent replicate experiments. Proteins that were significantly enriched with TRF1-FokI were those with an average LFC of >1. REACTOME pathway enrichment analysis was performed on the significant proteins using the clusterProfiler Bioconductor package (version 4.6.2)^[224]71 and the ReactomePA Bioconductor package (version 1.42.0)^[225]72. Measuring cell cycle Cell cycle analysis using flow cytometry was performed as previously described^[226]65. Cells were harvested and subsequently fixed with a 70% ethanol solution. After fixation, the cells underwent a single wash in PBS before being resuspended in PI/RNase Staining Buffer (Invitrogen). Following a 20-minute incubation at 37 °C, the stained cells were subjected to analysis via a flow cytometer (BD Biosciences). The cell cycle distributions were analyzed by ModFit LT™. AlphaFold2-Multimer structure prediction Protein sequences for CHAMP1 (accession# [227]Q96JM3), POGZ (accession# [228]Q7Z3K3), HP1α (accession# [229]P45973) and SETDB1 (accession# [230]Q15047) were retrieved from UniProtKB. Initially, full-length structural models of CHAMP1-POGZ, HP1α-POGZ, and POGZ-SETDB1 complexes were predicted using a locally installed ColabFold^[231]42,[232]73,[233]74, on a GPU machine. For each prediction, alphafold2 (AF2)_multimer_v3 with the default parameters for the run including num_recycles = 20, num_models = 5 and amber relaxation were used. After analyzing the resulting structural models of the full-length protein complexes for the most likely interface from each pair, a more focused prediction was performed using truncated protein sequences involving only protein domains that were part of the interface. For CHAMP1-POGZ complex structure prediction, CHAMP1 N-terminus Zn finger domain (1–87 aa) and POGZ C-terminus domain (1021–1410 aa) sequences were used as input. For CHAMP1-HP1α complex structure prediction, CHAMP1 C-terminus Zn finger domain (694–812 aa) and HP1α N-terminus domain (1–80 aa) sequences were used as input. For POGZ-SETDB1 complex structure prediction, POGZ Zn finger core (468–693 aa) and SETDB1 sequence (560–1291 aa) that incorporates Methyl-CpG-binding domain and two SET domains from the C-terminus were used as input. The AF2-mutimer-generated structural models were reproduced by recently released AF3 server ([234]https://golgi.sandbox.google.com/) using the same input sequences. For HP1α-POGZ model using AF3 server, we used two copies of HP1α C-terminus domain (109–180 aa) (as HP1α homodimerizes using the C-terminus based on the crystal structure PDB:3I3C) and two copies of POGZ HPZ-domain (791–850 aa) (as POGZ was shown to bind to HP1α using its HPZ-domain^[235]19) sequences were used as input. For each prediction, the resulting models were ranked based on the interface-predicted template modeling (iptm) score. Only the top models from each run were used for further analyses. All structural analyses and figures were generated using Pymol (Schrödinger, Inc). All AF2 models were experimentally verified by the immunoprecipitation protocol described above. Quantification and statistical analysis All values are expressed as mean ± standard deviation or standard error of the mean as indicated in figure legends. The statistical significance of differences was assessed by Student’s t test for comparison of two groups, one-way analyses of variance with Tukey’s test for comparison of multiple groups using GraphPad Prism 10. The microscopy-related images, including those from immunofluorescence and IF-FISH, were acquired randomly. Reporting summary Further information on research design is available in the [236]Nature Portfolio Reporting Summary linked to this article. Supplementary information [237]Supplementary Information^ (31.8MB, pdf) [238]Reporting Summary^ (307.4KB, pdf) [239]Transparent Peer Review file^ (784KB, pdf) [240]Supplementary Data 1^ (367.3KB, xlsx) [241]Supplementary Data 2^ (392.6KB, xlsx) Source data [242]Source Data^ (97.5MB, zip) Acknowledgements