Dynamic chromatin tethering of MDC1 regulates genome stability Authors: Joshua R Heyza, Mariia Mikhova, Cody Phillips, Shomaila Mehmood, Gloria I Perez, David G Broadbent, Jens C Schmidt Published: 2026 ABSTRACT AbstractDNA double-strand breaks (DSBs) are highly cytotoxic DNA lesions because they disrupt the connectivity of the DNA strand. Homologous recombination (HR) is a high-fidelity DSB repair pathway that copies the sequence spanning the break from a homologous template, but how DNA ends are held together during HR remains unclear. Here we demonstrate that the proline-serine-threonine (PST) repeat region of Mediator of DNA Damage Checkpoint 1 (MDC1) is a multivalent nucleosome-binding domain, sufficient to tether chromatin in multiple contexts. In interphase, the constitutive chromatin association of MDC1 is critical for RAD51 loading and efficient HR. In mitosis, PST-mediated chromatin binding is attenuated by phosphorylation, preventing aberrant chromosomal interactions while preserving DNA tethering by the MDC1–TOPBP1–CIP2A complex. In total, this work demonstrates that the PST repeat region of MDC1 is a multivalent nucleosome-binding domain with tunable affinity that supports DSB repair by HR and maintains genome stability during mitosis. Introduction DNA double-strand breaks (DSBs) are widely considered to be the most toxic genomic lesion. DSBs can arise in all cell-cycle phases and are induced both as part of intrinsic cellular processes and as a consequence of exposure to endogenous and exogenous chemicals [1]. These lesions pose significant threats to genome stability because the failure to repair DSBs can lead to loss of genetic information, insertions and deletions, or translocations, which can result in the development of a wide array of diseases, including immunodeficiencies and cancers [2–4]. The cellular response to DSBs is orchestrated by the phosphatidylinositol 3-kinase-related kinases DNA-PKcs, ATM, and ATR, which initiate a chromatin-based signaling cascade at sites of damage [5]. A key early event in this response is phosphorylation of histone H2AX at serine 139 (γH2AX), which recruits the adaptor protein Mediator of DNA Damage Checkpoint 1 (MDC1) through its C-terminal BRCT domains [5–7]. MDC1 is a large, multifunctional protein that coordinates downstream signaling and repair factor recruitment [8]. At DSBs in interphase, MDC1 promotes high-fidelity homologous recombination (HR) and error-prone non-homologous end joining (NHEJ) repair by facilitating the recruitment of BRCA1 and 53BP1 to DNA breaks [6, 7, 9–14]. This is accomplished by a direct interaction between MDC1 and the E3 ubiquitin ligase, RNF8, through a cluster of TQXF motifs in MDC1 that are phosphorylated by ATM [12]. This interaction initiates RNF8-mediated poly-ubiquitination of histone H1 that creates a docking site for the E3 ubiquitin ligase, RNF168, which enables BRCA1 and 53BP1 localization to DSBs by depositing ubiquitin moieties on H2A at K13 and K15 [14–17]. BRCA1 and 53BP1, in turn, carry out critical functions in regulating DNA end resection and the balance between HR and NHEJ pathways [18–25]. Mitotic cells are particularly vulnerable to DSBs, since they can lead to loss of genetic information and/or ploidy changes [26]. HR and 53BP1-dependent end joining are suppressed during cell division by inhibiting the RNF8-MDC1 interaction via CDK1-mediated phosphorylation of RNF8 at T198 and by phosphorylation of 53BP1 [27]. Recent work demonstrated that alternative end joining mediated by polymerase theta is activated in mitosis to enable the repair of DSBs [28, 29]. Additionally, MDC1 binds to mitotic DSBs marked by γH2AX, where phosphorylation of MDC1 at S168 and S196 by CK2 facilitates an interaction with TOPBP1 and the formation of the MDC1–TOPBP1–CIP2A complex. Together the MDC1–TOPBP1–CIP2A complex physically tethers broken DNA ends together through mitosis and into G1-phase, where they are repaired by 53BP1-dependent end joining [30, 31]. The MDC1–TOPBP1–CIP2A axis has also been implicated in tethering of shattered micronuclear chromosome fragments, ensuring their asymmetric inheritance to a single daughter cell [64, 65]. In addition to several well-characterized domains and protein-protein interaction sites, MDC1 contains a large intrinsically disordered region known as the proline-serine-threonine (PST) repeat region, whose function remains poorly understood. The PST repeat region, which is unique to mammals, is defined by a 41-amino-acid sequence, with 13 repeats present in human MDC1 [32, 33]. Previous studies have reported interactions between the PST repeat region and multiple factors, including DNA-PKcs, Chk1, FBXW7, Rag1, and pre-rRNA, and have linked this domain to checkpoint signaling and protein stability [34–37]. More recently, our work and that of others has demonstrated that the PST repeat region mediates constitutive, DNA damage-independent chromatin association of MDC1 in interphase through interaction with the nucleosome acidic patch [38, 39]. In this study, we demonstrate that the PST repeat region of MDC1 functions as a multivalent chromatin-binding domain that is sufficient to tether DNA. We show that this constitutive chromatin association is dynamically regulated during mitosis through CDK1-dependent phosphorylation of multiple threonine-proline (TP) motifs within the PST repeats. Using custom phospho-specific antibodies, we demonstrate that PST phosphorylation occurs upon nuclear envelope breakdown and that phosphorylated MDC1 persists at mitotic DSB sites into early G1 phase. While this phosphorylation does not impact MDC1 focus formation or assembly of the MDC1–TOPBP1–CIP2A complex, we find that loss of the PST repeat region disrupts MDC1 focus formation in mitosis. Conversely, preventing phosphorylation of PST repeats results in aberrant accumulation of MDC1 on mitotic chromatin, leading to abnormal chromatin bridging between condensed mitotic chromosomes and impairing their separation during anaphase. In interphase the PST repeat region is dispensable for MDC1 focus formation and 53BP1 recruitment but is required for RAD51 loading and efficient HR. Together, these findings suggest that the PST repeat region functions as a chromatin-binding module that contributes to MDC1-dependent DNA repair across the cell cycle.brc Materials and methods Cell lines and cell culture U2OS, RPE1 p53+/+, and HEK293T cells were obtained from ATCC and cultured in RPMI (U2OS and RPE1) or Dulbecco’s modified Eagle’s medium (HEK293T) containing 10% fetal bovine serum, and 100 units/ml penicillin, and 100 µg/ml streptomycin in a humidified incubator maintained at 37°C with 5% CO2. 3×FLAG-Halo-MDC1 knock-in and ΔMDC1 knockout U2OS cells were previously described in Heyza et al. [38]. DR-GFP knock-in into the AAVS1 locus of ΔMDC1 U2OS cells was previously described in Heyza et al. [38]. U2OS 2-6-3 cells containing the Lac operator (LacO) array were a kind gift from David Spector and were maintained in medium containing 100 µg/ml hygromycin B [40]. Molecular cloning and plasmids DNA oligonucleotides were purchased from IDT for cloning guide RNA (gRNA) sequences. After annealing, gRNAs were cloned into BpiI-digested pX330-U6-Chimeric_BB-CBh-hSpCas9 [a gift from Feng Zhang (Addgene plasmid # 42230)] [41]. The gRNA/Cas9 and homology-directed repair donor plasmids used to introduce a 3×FLAG-Puro-HaloTag into the endogenous MDC1 locus in RPE1 cells were previously described in Heyza et al. [38]. Sequences and cloning of Halo-MDC1 wild-type (WT), ΔPST, and ΔBRCT cloned into pRK2 (modified from pHTN HaloTag CMV-neo; Promega; #G7721) by Gibson assembly were previously described in Heyza et al. [38]. MDC1 T > D and T > A mutants were cloned into pRK2 using N- and C-terminal fragments flanking the PST repeat domain that were polymerase chain reaction (PCR) amplified from WT MDC1 and two codon-optimized gene fragments encoding the PST domain for each mutant, which were purchased as gBlocks from IDT. To generate PST truncation mutants, we first cloned into pRK2 a WT version of MDC1 containing a codon-optimized PST repeat region using the same approach used to generate the T > D and T > A mutants. Next, we PCR amplified the N-terminus of MDC1 and generated repeat truncations by amplifying the C-terminus using staggered forward primers in PST domain. N- and C-terminal fragments were cloned by Gibson assembly into pRK2 after linearization by inverse PCR. For tetracycline-inducible expression of MDC1 mutants, MDC1 complementary DNAs (cDNAs) were amplified from pRK2 and cloned by Gibson assembly into pSBTet-Bla after linearization with SfiI [a gift from Eric Kowarz (Addgene plasmid # 60510)] [42]. The Sleeping Beauty transposase expression vector, SB100×, was a gift from Mark Groudine (Addgene plasmid # 127909) [43]. To knock in the ΔPST mutant into the MDC1 locus, we generated a homology directed repair (HDR) donor plasmid by Gibson assembly containing left and right homology arms (one ordered as a gBlock and the other PCR amplified from MDC1 cDNA), C-terminal MDC1 cDNA (PCR amplified), and a puromycin resistance cassette (PCR amplified). To knock in the WT, T > A, and T > D PST mutants in 3×FLAG-Halo-MDC1 U2OS and RPE1 cells, we generated an HDR plasmid by Gibson assembly containing left and right homology arms and a codon-optimized PST repeat sequence (ordered as two gBlocks). The H2A K13/K15 ubiquitin sensor expression plasmid, NLS-Reader1.0-eGFP, was a kind gift from Robert Cohen and Tingting Yao [44]. For lentiviral expression, the NLS-Reader1.0-eGFP was PCR amplified and cloned by Gibson assembly into AgeI- and EcoR1-linearized FUGW [a gift from David Baltimore (Addgene plasmid # 14883)] [45]. To knock in inducible WT and T > A GFP-PST-LacI fusion proteins into the AAVS1 locus in U2OS 2-6-3 cells, we PCR amplified GFP and LacI from a GFP-LacI expression plasmid (a kind gift from Iain Cheeseman) [46] and codon-optimized WT and T > A PST repeat sequences and cloned them by Gibson assembly into MluI- and SalI-linearized AAVS1-TRE3G-EGFP [a gift from Su-Chun Zhang (Addgene plasmid # 52343)] [47]. The AAVS1-targeting gRNA plasmid, PX458-AAVS1, was a gift from Adam Karpf (Addgene plasmid # 113194). To express and purify GST-tagged PST repeats from Escherichia coli, two gBlocks for each protein fragment were codon-optimized for expression in E. coli and were cloned by Gibson assembly into BamHI-linearized pGSTag (a gift from David Ron, (Addgene plasmid # 21877) [48]. All newly generated plasmids were validated by Sanger sequencing or next-generation sequencing (Quintara Biosciences and Genewiz). Lentivirus Lentiviral particles to express NLS-Reader1.0-eGFP were produced in HEK293T cells. HEK293T cells were seeded onto 10-cm plates until they reached ∼70% confluency. Cells were co-transfected with FUGW (NLS-Reader1.0-eGFP), pRSV-Rev, pMDLg/pRRE, and pMD2.G with FuGene6 at a 3:1 ratio of FuGene6:DNA in antibiotic-free medium and incubated overnight. The following day, transfection medium was removed, and 10 ml of antibiotic-free medium was added. Virus was collected at 48 and 72 h post-transfection, filtered through a 0.4-µm filter, and 1-ml aliquots frozen at −80°C. For lentiviral transduction, cells were seeded onto 10-cm plates and allowed to attach overnight. The following day, viral aliquots were warmed to 37°C, 8 µg/ml polybrene was added, and virus was added dropwise to each plate. Experiments were carried out with the pooled population of transduced cells. Genome editing To knock in the ΔPST, T > D, T > A, and WT PST repeats in the endogenous Halo-Tagged MDC1 locus, GFP-PST-LacI into the AAVS1 locus, and 3×FLAG-Puro-HaloTag into the endogenous MDC1 locus in RPE1 cells, Halo-MDC1 U2OS, U2OS 2-6-3, or RPE1 cells were transfected with 1 µg each of pX330 and the HDR donor overnight using FuGene6 at a 3:1 ratio of Fugene6:DNA in antibiotic-free medium. The following day, transfection medium was replaced with complete medium. Knock-in cells were selected for with 1 µg/ml (PST) or 1.5 µg/ml puromycin. For WT and T > A GFP-PST-LacI, a single knock-in clone survived selection. ΔPST knock-in clones were isolated in 96-well plates by single-cell dilution, screened using genomic PCR, and validated by in-gel fluorescent protein detection and Sanger sequencing. RPE1 Halo-MDC1 cells were transfected with a plasmid encoding GFP-Cre to recombine out the puromycin resistance cassette oriented between the 3×FLAG and HaloTag and sorted by fluorescence-activated cell sorting into 96-well plates. Clones were screened by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and genomic PCR. Phospho-PST antibody production Sera containing phospho-specific antibodies were generated against the following peptides (Ac-DQPV (pT)PEPTSC-amide, Ac-CNRSSVK (pT)PE-amide) by Labcorp using two rabbits per target and a 118-day protocol. All bleeds from both rabbits were pooled and purified. Peptides containing the phosphorylated and unphosphorylated peptides were crosslinked to SulfoLinkTM colums (Thermo Scientific, 44999). Sera were first flowed over columns with the unphosphorylated peptide to deplete antibodies that recognized parts of the peptide not specific to the phosphorylated threonine. The flowthroughs from these columns were flown over the columns with cross-linked phosphorylated peptide, washed, and eluted using acidic and basic conditions. Fractions containing high protein concentrations from both elution conditions were pooled and dialyzed into phosphate-buffered saline (PBS) containing 50% glycerol, spun down at 20 000 × g for 15 min, and supernatants were aliquoted, snap-frozen in liquid nitrogen, and stored at −80°C. SDS–PAGE and western blot Cells grown in 24-well or 6-well plates were used for in-gel fluorescence and/or western blot analysis. Where indicated, cells were arrested in mitosis by treatment with 150 ng/ml nocodazole overnight. HaloTag labeling was performed by incubating cells with 100 nM Halo-JF646 or JFX650 for 10 min at 37°C, followed by washing with complete medium or PBS to remove unbound ligand. For in-gel fluorescence, cells were washed once with PBS and lysed directly in wells using 2× Laemmli buffer (Bio-Rad; #1610737) supplemented with β-mercaptoethanol. For western blot analysis, cells were lysed in RIPA buffer (Thermo Scientific, #89900) with protease and phosphatase inhibitors (HaltTM Protease and Phosphatase Inhibitor cocktail, Thermo Scientific, #78440) for protein quantification using a BCA protein assay kit (Thermo Scientific, #A55865), and 40 µg of protein per sample was mixed with 2× Laemmli buffer containing β-mercaptoethanol. All samples were boiled at 95°C for 5 min prior to loading onto 4%–15% or 4%–20% TGX stain-free polyacrylamide gels (Bio-Rad). In-gel fluorescence of Halo-MDC1 was detected using a Cy5.5 filter on a Bio-Rad ChemiDoc imaging system. Total protein loading was assessed using stain-free detection following 45 s activation with UV. For western blot analysis, proteins were transferred onto nitrocellulose membranes (Cytiva, #10600003). Low molecular weight proteins (e.g. H3 pS10) were transferred using a TransBlot Turbo semi-dry transfer system (Bio-Rad), and high molecular weight proteins (e.g. Halo-MDC1) were transferred overnight with CAPS buffer (10 mM CAPS, 10% methanol, pH ∼10.5–11). Membranes were blocked in 5% milk in 0.05% PBS-Tween 20 (PBS-T) for 1 h at room temperature, washed three times with PBS-T, and incubated with primary antibodies against phospho-histone H3 (Ser10) (Cell Signaling Technology, #9701S, 1:1000) and pPST1 (1:1000) in PBS-T for 1 h at room temperature. Membranes were then washed three times with PBS-T and incubated with HRP-conjugated secondary antibody (Cytiva, #NA934V, anti-rabbit IgG; 1:2000 in 5% milk) for 2 h at room temperature. Blots were developed using enhanced chemiluminescence substrate (SuperSignal West Femto Max Sensitivity Substrate, Thermo Scientific, #34096) and imaged on a Bio-Rad ChemiDoc system. Colony survival assay Six hundred to 850 cells/well were plated in six-well plates in complete medium and allowed to attach overnight. The following day, medium was removed, and fresh medium containing increasing concentrations of olaparib (Selleckchem; S1060) or zeocin was added for the duration of the experiment. Approximately 7 days after treatment, medium was removed, wells washed once with PBS, and cells fixed/stained with crystal violet solution (20% ethanol, 1% w/v crystal violet). After removing staining solution, plates were washed with water and allowed to air-dry. Plates were imaged on a Bio-Rad ChemiDoc, and colonies counted using OpenCFU [49]. Experiments were performed three times in triplicate. Timelapse imaging of MDC1 Timelapse imaging was carried out on a Yokogawa CellVoyager CQ1 Benchtop High-Content Analysis System using a 40× objective. Tet-inducible Halo-MDC1 cells were seeded onto 24-well glass-bottom plates 2 days prior to imaging and MDC1 expression was induced with 2 µg/ml doxycycline. Cells were maintained at 37°C with 5% CO2 during the course of imaging. Multiple fields of view were randomly selected for each sample and 3D images were acquired every 5 min for ∼20 h. Z-projected images were analyzed in ImageJ/Fiji, where the mean MDC1 fluorescence intensity for each analyzed cell was measured prior to mitotic entry, and time to anaphase onset was measured by manually counting the number of frames starting with the first frame where dissolution of the nuclear envelope was observed until chromosome segregation or attempted segregation was initiated. At least 100 cells were analyzed for each mutant. Timelapse imaging of Halo-MDC1, ΔMDC1, and PST mutant knock-in cell lines (WT, ΔPST, T > D, and T > A) was carried out on a Sartorius Incucyte SX 3 using a 20× objective. Cells were maintained at 37°C with 5% CO2 during the course of imaging. Brightfield images were acquired every 5 min for ∼3 days for mitotic timelapses or every 2 h for monitoring growth rates. Mitotic timelapse images were analyzed in ImageJ/Fiji (200 cells were analyzed for each group). Microscopy Experiments were carried out on four microscope systems. The first system was a Olympus IX83 inverted microscope equipped with an environmental control unit, a cellTIRF illuminator (405, 488, 561, and 641 nm laser lines), an Excelitas X-Cite TURBO LED light source, Olympus UAPO 100× (1.49 NA) and 60× (1.5 NA) TIRF objectives, a cellFRAP with a 100 mW 405 nm laser, and 2 Andor iXon 897 Ultra EMCCD, Hamamatsu Orca-Fusion BT sCMOS, or Hamamatsu Orca-Quest qCMOS cameras, and operated using Olympus cellSense software. The second system was a 3i spinning-disc confocal (Yokogawa) microscope with SoRa modality equipped with incubation chamber (temperature, humidity, and CO2 controlled), four laser lines (445, 488, 561 and 638 nm), Zeiss C PlanApo 63×/1.42 NA objective and Hamamatsu Orca-Fusion BT sCMOS or Hamamatsu Orca-Quest qCMOS camera. The third system was a Zeiss LSM800 confocal laser scanning microscope equipped with four laser lines (405, 488, 561, and 639 nm) and a Zeiss C PlanApo 63×/1.42 NA objective housed in the Wayne State University Microscopy, Imaging, and Cytometry Resources core facility. The fourth system was an Abbelight SAFe M45 equipped with equipped with an environmental control unit, four laser lines (405, 488, 561, and 638 nm laser lines), Olympus UAPO 100× (1.50 NA) objective, and Hamamatsu Orca-Fusion BT sCMOS camera. Live-cell imaging All live-cell imaging experiments were carried out at 37°C, 5% CO2, and 95% humidity-controlled environment to mimic the conditions used for cell culturing. Cells were seeded into 24-well glass-bottom plates for live-cell and live-cell single-molecule imaging. For imaging of densely labeled Halo-MDC1 mutants in interphase and mitosis, ΔMDC1 cells were transfected with plasmids encoding Halo-WT, T > A, T > D, ΔPST, or ΔBRCT MDC1 in six-well plates 2 days prior to seeding for imaging using FuGene6. Alternatively, knock-in cell lines expressing different PST mutants were directly seeded onto 24-well glass-bottom plates on the day prior to imaging. The following day, 100 ng/ml nocodazole was added to each well, and cells were incubated for ∼16 h prior to imaging in mitosis. Immediately prior to imaging, Halo-MDC1 and DNA were labeled with 100–250 nM Halo-JF646 or Halo-JFX650 and Hoechst for ∼10 min (in the presence of nocodazole for mitotic samples), washed three times with complete medium, and fresh medium (containing nocodazole for mitotic samples) was added. For live-cell single-molecule imaging of Halo-MDC1 mutants in interphase and mitosis, cells were transfected using FuGene6 with plasmids encoding each mutant in six-well plates and subsequently seeded for imaging or were nucleofected (PST truncation mutants) using a Lonza 4D Nucleofector and plated onto 24-well glass-bottom plates. For single-molecule imaging in mitosis, samples were plated onto 24-well glass bottom plates and allowed to attach overnight. Approximately 24 h later, 100 ng/ml nocodazole was added to each well, and cells were incubated for ∼16–20 h prior to imaging. Immediately prior to imaging, Halo-MDC1 was labeled with either Halo-JF646 or Halo-JFX650 (0.5–4 nM for 1 min). Excess ligand was removed by washing the samples three times with complete medium; incubated for 10 min at 37°C, and fresh complete medium added. Imaging was done using either the 100× or 60× objective and the 640 nm laser line with highly inclined laminated optical sheet illumination. Images were acquired at 138–198 fps for at least 1000 frames. Single-molecule imaging analysis Live-cell single-molecule imaging movies were analyzed using single-particle tracking (SPT) in MATLAB 2019a with a version of SLIMfast that allows for analysis of TIFF files [50]. SPT setting use for analysis: Exposure Time = 5–7.20 ms, NA = 1.49 or 1.5, Pixel Size = 0.16 mm (Andor iXon Ultra) or 0.1083 mm (Hamamtsu ORCA Fusion BT), Emission Wavelength = 664 nm, Dmax = 5 mm2/s, Number of gaps allowed = 2, Localization Error = −5, Deflation Loops = 0. Next, particle tracks were analyzed using SpotOn in MATLAB to obtain diffusion coefficients and the fraction of bound and freely diffusing particles. SpotOn settings used for the analysis of SPT files are as follows: TimeGap = 5–7.20 ms. dZ = 0.700 mm, GapsAllowed = 2, TimePoints = 8, JumpsToConsider = 4, BinWidth = 0.01 mm, PDF-fitting, D_Free_2State = [0.5 25], D_Bound_2State = [0.0001 0.5]. At least 20 cells for each of three biological replicates were imaged for all live-cell single-molecule imaging experiments. Summary data for each biological replicate were plotted in GraphPad Prism. Laser microirradiation Laser microirradiation (LMI) experiments were carried out using the Olympus ix83 microscope described earlier. Cells were seeded onto 24-well glass-bottom plates for LMI experiments. For tet-inducible MDC1 mutants, 2 µg/ml doxycycline was added 1–2 days prior to imaging to induce protein expression. For dual-color imaging of Halo-MDC1 and NLS-Reader1.0-eGFP, cells were transduced with NLS-Reader1.0-eGFP lentiviral particles in 10-cm plates, were subsequently seeded onto 24-well glass-bottom plates, and Halo-MDC1 expression was induced with doxycycline. For LMI carried out on mitotic cells, mitotic cells were enriched by adding 100 ng/ml nocodazole to each well for ∼16–20 h prior to imaging. On the day of imaging, Halo-MDC1 was labeled with 100–150 nM Halo-JFX650 and simultaneously pre-sensitized with Hoechst (1 µg/ml) for 10 min (in the presence of nocodazole for mitotic samples). After labeling, cells were washed three times with complete medium and incubated for 10 min at 37°C, followed by the addition of fresh complete medium. Cells were maintained at 37°C with 5% CO2 during the course of imaging. Cells were irradiated using an interpolated line at 20% laser power with a 20 ms pulse. Images were acquired using a 60× objective, and fluorescence imaged after excitement with the 475 or 630 nm LED light source at various rates (e.g. every 0.5, 1, and 5 s). Images were converted to TIFF files, and Halo-MDC1 and NLS-Reader1.0-GFP recruitment were quantified in Fiji by placing an region of interest (ROI) around the irradiated line and measuring mean intensity over time. From these data, the mean fluorescence at baseline was subtracted to calculate the relative increase in fluorescence over time. Normalized fluorescence intensity values were determined by setting the highest intensity frame for each cell to 1. Similarly, for normalizing the intensity values for the average of all cells combined for each group, the frame with the highest mean intensity was set to one. Data were plotted in GraphPad Prism. Purification of GST-PST repeats and in vitro kinase assay Plasmids encoding WT and T > A GST-PST repeats were transformed into OneShotTM BL21 (DE3) cells (Invitrogen). Bacterial cultures were grown at 37°C at 220 rpm to an OD600 of 0.7. Protein expression was induced with 1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) and cultures grown for ∼16 h at 18°C with shaking at 180 rpm. Cells were harvested by centrifugation at 5000 × g for 5 min at 4°C and frozen at −80°C. Pellets were resuspended in wash and lysis buffer [250 mM NaCl, 1 mM dithiothreitol (DTT) (added fresh) in PBS]. Lysozyme was added at a concentration of 0.5 mg/ml, and samples were sonicated (90 s, 35% amplitude, 10 s pulses, 20 s pause, Fisherbrand Model 505, 0.5-inch tip) in an ice-water bath. Lysates were cleared by centrifugation at 40 000 × g for 30 min at 4°C. Two milliliters of Glutathione Sepharose (GE Healthcare; 17-5132-01) was equilibrated by washing three times with 20 ml of wash and lysis buffer and spun down at 1000 × g for 2 min between washes. After clearing, lysate was added to the GST agarose and rotated for 1 h at 4°C. Samples were spun down at 1000 × g for 2 min, lysate removed, and samples were washed 3× with 10 ml of wash and lysis buffer. After washing, 4 ml of elution buffer [50 mM Tris, pH 8.1, 75 mM KCl, 10 mM glutathione (added fresh)] was added to each sample, rotated for 10 min at 4°C, and eluted protein collected. GST-PST repeats were additionally purified using size-exclusion chromatography using a Superdex 200 column into 50 mM Tris, pH 7.0, 150 mM KCl, 1 mM DTT. Fractions were run on an SDS–PAGE gel, and peak fractions were combined and concentrated. Concentrated protein was supplemented with 50% glycerol, frozen in liquid nitrogen, and stored at −80°C. Protein concentration was measured by absorption spectroscopy using ɛ280 nm = 46 090 M−1 cm−1, which was calculated using the GST-PST sequence and ExPASy ProtParam. In vitro kinase assays with purified Cyclin B1/CDK1 and WT/T > A GST-PST repeats were performed following a protocol described in Wang et al. [51]. Purified human Cyclin B/CDK1 was purchased from Invitrogen (PV3292). 2.5 µg GST-PST and ∼380 ng Cyclin B/CDK1 were incubated in 1× kinase buffer (0.25 M Tris–HCl, pH 7.5, 0.05 M MgCl2, 0.5 M ethylenediaminetetraacetic acid, 10 mM DTT), supplemented with 100 µCi/mmol of [γ-32P] ATP (PerkinElmer) and incubated at 30°C for 1 h. The reaction was stopped by adding 2× Laemmli buffer containing β-mercaptoethanol. Samples were incubated at 60°C for 10 min then loaded onto SDS–PAGE gels. Samples were visualized using autoradiography with an Amersham Typhoon (GE). Flow cytometry DR-GFP assays were performed three to four times, and fluorescence was detected by flow cytometry. ΔMDC1 DR-GFP knock-in cells were nucleofected using a Lonza 4D nucleofector in RPMI plus 50 mM HEPES with 400 ng I-SceI plasmid [a gift from Maria Jasin (Addgene plasmid # 26477)] [52] and 600 ng of plasmids encoding Halo-MDC1 mutants or 3×FLAG-Halo-NLS (previously described Klump et al. [53]). Forty-eight hours after nucleofection, HaloTag was labeled with 150 nM Halo-JFX650, and cells were trypsinized and collected. Samples were run on a BD Accuri C6 cytometer, and at least 5000 JFX650-positive events collected for each sample and analyzed in FCS Express 7. Cell cycle analyses data were acquired on a 3-laser Cytek Northern Lights (Cytek Biosciences, San Jose, CA) at the Wayne State University Microscopy, Imaging, and Cytometry Resources core. Data were analyzed using FlowJo (v10.8, Waters Biosciences, Ashland, OR) software. Immunofluorescence For immunofluorescence (IF), cells were seeded onto glass coverslips in six-well plates and allowed to attach for 1 day. For PST truncation mutants, ΔMDC1 cells were nucleofected with each Halo-MDC1 plasmid using a Lonza 4D nucleofector in serum- and antibiotic-free RPMI medium supplemented with 50 mM HEPES buffer solution (GIBCO; 15630-106) immediately prior to seeding. For tet-inducible MDC1, protein expression was induced by adding 2 µg/ml doxycycline for 2 days prior to proceeding with IF. For mitotic samples, cells were synchronized at G2/M with CDK1i (9 µM; Selleckchem; S7447) for ∼16–20 h. HaloTagged protein was labeled at the time of release from CDK1i by adding 100 nM Halo-JFX650 and incubating for 10 min. Samples were washed three times to remove unbound ligand and released into medium containing S-Trityl-L-cysteine (10 µM; Enzo Life Sciences, ALX-105-011-M500) and ±zeocin (20 µg/ml; Invitrogen; R250011) for 1 h. For images of different stages of mitosis, samples were released from CDK1i into complete medium ± zeocin and fixed at 60, 90, or 120 min post-release. For interphase samples, HaloTagged protein was labeled with 100 nM Halo-JFX650 for 10 min, followed by three washes with complete medium and subsequent incubated for 4 h ± zeocin (20 µg/ml). Cells were washed once with PBS and fixed with 4% formaldehyde or ice-cold 100% methanol (for TOPBP1 or p-PST staining) for 10 min at room temperature. Fixative was removed by washing samples once with PBS, followed by permeabilization with 0.2% Triton-X 100 in PBS. Samples were washed twice with antibody dilution buffer (ABDIL) [3% bovine serum albumin in PBS-Tween 20 (PBS-T)] and blocked in ABDIL for 30–60 min. Samples were incubated with primary antibodies for 1–1.5 h at room temperature, washed three times with PBS-T, and incubated with secondary antibodies for 1 h at room temperature. After washing three times with PBS-T, DNA was labeled with Hoechst, coverslips washed once with PBS, mounted onto glass slides with ProLong Diamond Antifade Mountant (Invitrogen; P36970), and sealed with nail polish. Antibodies: anti-γH2AX (Millipore; JBW301; 05-636) (1:1, 000); anti-TOPBP1 (Bethyl; A300-111A) (1:500); anti-CIP2A (Santa Cruz; sc-80659) (1:500); anti-53BP1 (Novus; NB100-304) (1:1000); anti-MDC1 (Invitrogen; P2B11; 1:500); anti-p-PST #1 (Stock concentration = 3 mg/ml) (1:3000); anti-p-PST #2 (Stock concentration = 6.2 mg/ml) (1:6000); goat anti-mouse IgG Antibody, Cy3 conjugate (Millipore; AP124C) (1:500); Cy3 goat anti-rabbit IgG (H + L) (Invitrogen; A10520) (1:500); goat anti-mouse IgG (H + L) AF488 (Invitrogen; A32723) (1:500 for CIP2A; 1:1000 for γH2AX). Goat anti-rabbit IgG (H + L) (Invitrogen; A11034) (1:500). Image analysis was performed in ImageJ/Fiji. Foci quantification was performed with the ComDet v0.5.5 plugin. Imaging of LacI-PST-GFP fusions in U2OS 2-6-3 cells U2OS 2-6-3 cells expressing inducible WT and T > A GFP-PST-LacI knocked into the AAVS1 locus were seeded onto 24-well glass-bottom plates and allowed to attach overnight. GFP-PST-LacI expression was induced with 1 µg/ml doxycycline. Time-lapse images were acquired using a 63× objective on the 3i spinning disk described earlier. Cells were not synchronized, and fields of view were selected based upon the presence of mitotic cells. Z-stack images for each XY position were acquired every 3 min for 1.5 - 2 h. ≥10 mitotic cells for each mutant were imaged per replicate on three separate days. The cell division phenotypes were assessed by manually inspecting each cell undergoing anaphase and analyzing the segregation of the LacO array signal as well as the overall cell morphology after cell division. Failed cytokinesis was apparent by persistent connection bridging the two cell bodies or by opening of the cleavage furrow leading to a binucleated cell. Results MDC1 is displaced from chromatin during mitosis Using live-cell single-molecule imaging, we previously demonstrated that HaloTagged MDC1 expressed from its endogenous locus in U2OS cells is constitutively associated with chromatin in interphase cells (Fig. 1A and B) [38]. To characterize the properties of MDC1 in a second cell line, we introduced the HaloTag into the endogenous MDC1 locus in RPE1 cells (Fig. 1C and Supplementary Fig. S1A). During our single-molecule imaging experiments in U2OS cells, we observed rare cells that displayed high Halo-MDC1 mobility, which appeared to be undergoing cell division. To determine the dynamics of Halo-MDC1 diffusion during mitosis, we synchronized cells in M-phase using nocodazole and carried out live-cell single-molecule imaging using highly inclined laminated optical sheet microscopy (Fig. 1A, Movies S1–S6) [54, 55]. SPT and data analysis using the SPOT ON tool demonstrated that in mitotic U2OS and RPE1 cells, the fraction of static Halo-MDC1 molecules, which represents MDC1 bound to chromatin, was ∼20%, while ∼60% and ∼75% of Halo-MDC1 molecules were immobile in interphase U2OS and RPE1 cells, respectively (Fig. 1D and E; Supplementary Fig. S1B–D) [50]. In addition, the diffusion coefficients of both mobile and static Halo-MDC1 molecules were increased in mitotic compared to interphase cells in both U2OS and RPE1 cells (Fig. 1F–I). These observations demonstrate that MDC1 is displaced from chromatin when cells enter mitosis. Phosphorylation of the PST repeat region removes MDC1 from chromatin in mitosis. (A) Model of MDC1 recruitment to chromatin. Fluorescence gels of cell lysates from (B) U2OS and (C) RPE1 expressing HaloTagged MDC1 from its endogenous locus labeled with JFX650 HaloTag-ligand. Static fraction of Halo-MDC1 molecules in interphase (I) and mitosis (M) derived from SPT of MDC1 in (D) U2OS and (E) RPE1 cells (N = 3 biological replicates, ≥20 cells per replicate, Mean ± SD, two-tailed t-test). Diffusion coefficient of (F) mobile and (G) static Halo-MDC1 molecules in interphase (I) and mitosis (M) derived from SPT of MDC1 in U2OS cells (N = 3 biological replicates, ≥ 20 cells and > 10, 000 trajectories per replicate, Mean ± S.D., two-tailed t-test). Diffusion coefficient of (H) mobile and (I) static Halo-MDC1 molecules in interphase (I) and mitosis (M) derived from SPT of MDC1 in RPE1 cells (N = 3 biological replicates, ≥20 cells and >10 000 trajectories per replicate, Mean ± SD, two-tailed t-test). (J) Domain organization of MDC1, sequence alignment of the thirteen 41-amino-acid PST repeats, and consensus sequence surrounding the two conserved TP sites within the PST repeats. (K) Phosphorescence imaging and protein staining of and SDS–PAGE gel loaded with samples of the WT and non-phosphorylatable (T > A) GST-tagged PST repeat domain treated with CDK1-Cyclin B and ATP γ-32P. (L) Static fraction derived from SPT of transiently expressed Halo-MDC1 mutants in cells arrested in mitosis with 100 ng/ml nocodazole [N = 3 biological replicates, ≥20 cells and >10 000 trajectories per replicate, Mean ± SD, two-way analysis of variance (ANOVA) (*P < .05; **P < .01; ***P < .001)]. Phosphorylation of its PST repeat region by CDK1-Cyclin B displaces MDC1 from chromatin during mitosis To define the molecular mechanism by which MDC1 is displaced from chromatin during mitosis, we focused on the PST repeat region of MDC1, which mediates its constitutive chromatin association in interphase cells [38, 39]. The human PST-region is composed of 13 imperfect 41-amino acid repeats (Fig. 1J). Analysis of MDC1 using the PhosphoSite.org database revealed that the PST repeats contain two TP sites, which have been shown to be phosphorylated in multiple high-throughput phospho-proteomic studies (Supplementary Fig. S1E) [56–59]. TP dipeptides are consensus sites for the mitotic CDK1-Cyclin B [60]. To test whether CDK1-Cyclin B can phosphorylate the PST repeat region of MDC1, we purified the GST-tagged WT PST repeat region of MDC1 and a variant of the PST repeat region in which all threonine residues within the TP sites were mutated to alanine (T > A, Fig. 1K). After incubation with CDK1-Cyclin B and ATP γ-32P, the WT PST repeat region was robustly phosphorylated by CDK1-Cyclin B, while phosphorylation of the T > A variant was reduced >10-fold (Fig. 1K), demonstrating that the TP sites within the PST repeats can be targeted by CDK1-Cyclin B. To determine whether phosphorylation of the TP sites within the PST repeat region is sufficient to displace MDC1 from chromatin during mitosis, we transiently expressed non-phosphorylatable (T > A) and phospho-mimicking (T > D) HaloTagged and performed SPT in nocodazole-arrested mitotic cells (Fig. 1L and S1F, Movie S7). Compared to WT MDC1, ΔPST, and ΔBRCT which all shared similar levels of bound particles (WT Fbound = 21.7 ± 3.5%, ΔPST Fbound = 25.2 ± 1.0%, ΔBRCT Fbound = 23.9 ± 3.0%), the T > A mutant was significantly enriched on mitotic chromatin (Fbound = 42.7 ± 8.0%) and the T > D mutant displayed an intermediate phenotype (Fbound = 33.3 ± 4.1%, Fig. 1L, Movie S7). These results demonstrate that phosphorylation of the TP sites within the PST repeat region of MDC1 helps to displace MDC1 from mitotic chromosomes. The number of PST repeats and their phosphorylation control the constitutive chromatin association of MDC1 To determine how the copy number of PST repeats and the phosphorylation of the TP sites in the repeats regulate the constitutive chromatin association of MDC1, we transiently expressed HaloTagged MDC1 variants in MDC1 knock-out cells and carried out live-cell single-molecule imaging experiments (Fig. 2A and Supplementary Fig. S1F). In interphase cells, the fraction of chromatin-associated MDC1 gradually decreased with removal of PST repeats (Fig. 2B, Movies S8–S15). Similarly, the diffusion coefficients of mobile and static MDC1 molecules gradually increased when more PST repeats were removed from MDC1 (Fig. 2C and D). This suggests that the contribution of each PST repeat to chromatin binding is comparable, rather than a specific set or a minimum number of repeats mediating the association of MDC1 with nucleosomes. The number of the PST repeats and TP phosphorylation regulates the chromatin association of MDC1. (A) Fluorescence gel of cell lysates of U2OS ΔMDC1 cells transiently expressing HaloTagged (JFX650) WT MDC1 and truncations of the PST repeats. Single particle tracking analysis of HaloTagged WT MDC1 and truncations of the PST repeats in interphase U2OS ΔMDC1 cells including the (B) fraction of static particles and the diffusion coefficients of (C) mobile and (D) static MDC1 molecules (N = 3 biological replicates, ≥20 cells and > 10 000 trajectories per replicate, Mean ± S.D.). Single particle tracking analysis of HaloTagged WT MDC1, MDC1 with a deletion of the PST repeat region (ΔPST) and MDC1 variants with phospo-mimicking (T > D) and non-phosphorylatable (T > A) mutations in the TP sites within the PST repeat region in interphase U2OS ΔMDC1 cells including the (E) fraction of static particles and the diffusion coefficients of (F) mobile and (G) static MDC1 molecules (N = 3 biological replicates, ≥20 cells and >10 000 trajectories per replicate, Mean ± S.D., two-way ANOVA). The dynamic properties of non-phosphorylatable (T > A) MDC1 were indistinguishable from WT MDC1 (Fig. 2E–G, Movies S16–S19), suggesting that the TP sites in MDC1 are not phosphorylated in interphase cells. In contrast, MDC1 with phospho-mimicking (T > D) TP sites in the PST repeat region had a reduced fraction of chromatin-bound molecules and an increased diffusion coefficient of static MDC1 molecules (Fig. 2E–G). Both the reduction in the static fraction and increase in the diffusion coefficient of chromatin-associated MDC1 T > D molecules was intermediate between WT MDC1 and MDC1 with a complete deletion of the PST repeat region (Fig. 2E–G). This suggests that MDC1 with phospho-mimicking (T > D) TP sites in the PST repeat region has a reduced affinity for chromatin, but it retains some ability to bind to nucleosomes. In total, these observations demonstrate that the ability of MDC1 to associate with chromatin is regulated by the copy number of and phosphorylation state of the TP sites within the PST repeats. In addition, these observations are consistent with the PST repeat region forming multivalent chromatin interactions, simultaneously binding to several nucleosomes. Phosphorylated MDC1 accumulates in mitosis in human cells To examine mitotic phosphorylation of MDC1, we raised phospho-specific antibodies against phosphorylated peptides encompassing the consensus sequences of the two TP sites within the PST repeats (Fig. 1J). Both phospho-specific antibodies recognized Halo-MDC1 foci in zeocin-treated mitotic, but not interphase, cells (Fig. 3A and B). To validate antibody specificity, we knocked out endogenous MDC1 in U2OS cells and re-expressed Halo-tagged MDC1 variants from the doxycycline-inducible Sleeping Beauty transposon cassette (Supplementary Fig. S2A) [42]. In mitotic cells, the phospho-TP #1 antibody co-localized with WT Halo-MDC1 foci but showed no foci in cells expressing ΔPST or TP mutants (T > A or T > D), confirming specificity (Fig. 3C and Supplementary Fig. S2B). Importantly, Halo-MDC1 lacking the PST repeat region displayed a strong defect in mitotic focus formation while all other Halo-MDC1 variants tested (WT, T > A, T > D) formed foci efficiently (Supplementary Fig. S2B). We also detected p-PST site #1 by western blot exclusively in nocodazole-arrested U2OS cells expressing WT MDC1. The p-PST site #1 signal was absent in lysate from asynchronous cells or from cells expressing the ΔPST mutant (Supplementary Fig. S2C) further confirming antibody specificity. Phosphorylated MDC1 accumulates at DNA breaks in mitosis and early G1 phase. Images of fixed interphase and mitotic U2OS cells expressing Halo-MDC1 (JFX650) immunostained with (A) p-PST antibody #1 or (B) p-PST antibody #2 and Hoechst to label DNA. (C) Normalized mean intensity of p-PST antibody #1 and #2 signal in interphase and mitotic cells (N = 3; >40 cells per replicate; red bar indicates median; one-way ANOVA). (D) Normalized median ratio per cell of p-PST #1 or #2 intensity relative to MDC1 at MDC1 foci in zeocin-treated (20 µg/ml; 1 h) interphase and mitotic cells (N = 3, >40 cells per replicate; red bar indicates median; one-way ANOVA. (E) Normalized ratio of p-PST #1 or #2 relative to MDC1 in all analyzed MDC1 foci in zeocin-treated (20 µg/ml; 1 h) interphase and mitotic cells (N = 3, >40 cells per replicate; red bar indicates median; one-way ANOVA). Cells were immunostained with p-PST antibody #1 and Hoechst to label DNA. (F, G) Images of fixed U2OS cells expressing Halo-MDC1 (JFX650) at different stages of mitosis (D) untreated or (E) treated with zeocin (20 µg/ml, 1 h) to induce DNA breaks in various phases of mitosis and early G1-Phase. Cells were immuno-stained with p-PST antibody #1, a γH2AX antibody, and Hoechst to label DNA. To quantify PST phosphorylation across the cell cycle, we performed IF using pPST1, pPST2, and MDC1 antibodies in Halo-MDC1 U2OS cells. Mean phospho-signal intensity, measured within the nucleus in interphase cells or across the whole cell in mitosis, was increased by >7-fold in mitotic cells, indicating that PST phosphorylation is a global modification of MDC1 that is largely restricted to mitosis (Fig. 3A–C). To determine whether this phosphorylation can also occur at DSBs in interphase, we quantified pPST signal relative to MDC1 intensity at zeocin-induced foci in interphase and mitotic cells. The normalized intensity ratio of pMDC1 relative to MDC1 was robust in mitotic cells but near zero in interphase cells (Fig. 3D), indicating that PST phosphorylation does not occur at DNA breaks during interphase. Analysis of all individual foci across cells yielded similar results, with interphase pPST:MDC1 intensity ratios averaging ∼2% of those observed in mitosis (Fig. 3E). Together, these results demonstrate that PST phosphorylation is mitosis-specific and does not occur at DNA damage sites during interphase. To determine the timing of MDC1 phosphorylation throughout cell division, we analyzed fixed cells in different stages of mitosis. In control cells not exposed to a DNA-damaging agent, phosphorylated MDC1 was diffusely localized throughout the cell during mitosis (Fig. 3F). Upon anaphase onset, phosphorylated MDC1 was detected throughout cells and began to enrich on DNA (Fig. 3F). By telophase, phosphorylated MDC1 was strongly enriched on chromatin (Fig. 3F). This suggests that MDC1 dephosphorylation is initiated immediately upon anaphase onset, and MDC1 rapidly transitions to its constitutively chromatin-bound state upon mitotic exit. In cells treated with zeocin to induced DSBs, phosphorylated MDC1 was detected at DNA lesions marked by γH2AX starting upon chromosome condensation in prophase through early G1 phase (Fig. 3F). These results were confirmed with the phospho-specific antibody targeting the second TP site within the PST repeats (Supplementary Fig. S2D and E). In total, these observations reveal that MDC1 phosphorylation is tightly coupled to mitotic progression and persists at DNA damage sites through mitotic exit into early G1. The PST repeat region is required for MDC1 accumulation at mitotic DNA breaks At mitotic DSBs, MDC1 forms a complex with TOPBP1 and CIP2A to tether broken DNA ends together into early G1 phase, where they are repaired [30, 31]. To analyze the role of the PST repeat region in mitotic DNA break recognition and tethering, we genome edited the endogenous HaloTagged MDC1 locus in U2OS cells to express MDC1 lacking the PST repeat region (Fig. 4A and Supplementary Fig. S3A–C, clones 33 and 37 are referred to as MDC1 ΔPST C1 and C2 throughout the manuscript). Cells expressing Halo-MDC1 ΔPST grew at a similar rate as their parental cells, indicating that the deletion of the PST repeat region of MDC1 did not result in a proliferation defect (Supplementary Fig. S3D). In mitotic cells expressing WT MDC1, MDC1 and TOPBP1 were robustly recruited to DNA breaks marked by γH2AX (Fig. 4B–D). In contrast, both MDC1 and TOPBP1 failed to accumulate at DSBs marked by γH2AX in cells expressing MDC1 lacking the PST repeat region (Fig. 4B–D). TOPBP1 accumulated in large foci in mitotic cells expressing MDC1 ΔPST, but they did not overlap with γH2AX and were also present in control cells not treated with zeocin, consistent with previous observations (Fig. 4B), suggesting that these TOPBP1 foci are not related to DNA break repair mediated by γH2AX [30]. To further analyze the role of the PST repeat region in recruiting MDC1 to DNA lesions in mitosis, we carried out LMI experiments in cells arrested in M phase using nocodazole. WT Halo-MDC1 was readily recruited to LMI-induced DNA lesions in mitotic cells (Fig. 4E, Movie S20). In contrast, only small amounts of Halo-MDC1 ΔPST accumulated at these sites of DNA damage (Fig. 4E, Movies S21 and S22). To independently validate the requirement of the PST repeat region for MDC1 focus formation at mitotic DNA breaks, we generated ΔPST knock-in clones in RPE1 cells (Supplementary Fig. S3H and I). Genome editing was confirmed by genomic PCR and in-gel fluorescence detection of Halo-MDC1 (Supplementary Fig. S3H and I). In mitotic cells, WT Halo-MDC1 was robustly recruited to zeocin-induced DSBs, whereas Halo-MDC1 ΔPST displayed a strong defect in focus formation (Supplementary Fig. S3J). TOPBP1 efficiently accumulated at DSBs in cells expressing WT MDC1 but did not form DNA damage-induced foci in ΔPST cells (Supplementary Fig. S3J–L). Quantification of the number MDC1 and TOPBP1 foci per cell confirmed a significant reduction in focus formation in cells expressing MDC1 ΔPST compared to WT controls (Supplementary Fig. S3K and L). In summary, these results validate in an independent cell line model that the PST repeat region is required for efficient recruitment of MDC1 and TOPBP1 to mitotic DSBs. The PST repeat region of MDC1 is required for focus formation at DNA breaks in mitosis. (A) Fluorescence gels of cell lysates from U2OS expressing HaloTagged MDC1 or MDC ΔPST from its endogenous locus labeled with JFX650 HaloTag-ligand. (B) Images of fixed, mitotic U2OS cells expressing HaloTagged MDC1 or MDC ΔPST from its endogenous locus untreated or treated with zeocin (20 µg/ml, 1 h) to introduce DNA breaks. Cells were labeled with JFX650 and Hoechst and immunostained with antibodies targeting TOPBP1 and γH2AX. Quantification of the number of (C) MDC1 and (D) TOPBP1 foci of the experiment shown in panel (B) with three biological replicates (>100 cells per condition, median ± S.D., two-way ANOVA). (E) Quantification of the accumulation of knock-in HaloTagged MDC1 and MDC ΔPST at laser micro-irraditation-induced DNA damage in living mitotic U2OS cells (n ≥ 10 cells). To determine the functional consequences of PST repeat phosphorylation for the recruitment of TOPBP1 and CIP2A to DSBs in mitosis, we generated additional homozygous knock-in U2OS cell lines expressing Halo-MDC1 variants containing WT PST repeats (WT KI control), phospho-mimetic (T > D), or non-phosphorylatable (T > A) mutations from the endogenous MDC1 locus. Successful genome editing was confirmed by PCR analysis of genomic DNA from individual clones (Supplementary Fig. S3E) and by in-gel fluorescence detection of Halo-MDC1 confirming expression of the expected fusion proteins (Supplementary Fig. S3F). Analysis of cell proliferation revealed that cells expressing the T > A mutant exhibited a growth defect comparable to MDC1 knockout cells, whereas cells expressing the WT PST or T > D mutant proliferated similarly to the WT KI control (Supplementary Fig. S3G). Next, we analyzed whether PST phosphorylation impacted MDC1 recruitment or assembly of TOPBP1 and CIP2A at mitotic DSBs. In contrast to the ΔPST mutant, which phenocopied MDC1 knockout cells, the WT KI, T > A, and T > D mutants all formed robust MDC1 foci at mitotic DSBs and supported recruitment of TOPBP1 and CIP2A (Fig. 5A–D). These results indicated that while the PST repeat region is required for MDC1-dependent focus formation at mitotic DSBs, PST phosphorylation is dispensable for MDC1 recruitment and assembly of the MDC1–TOPBP1–CIP2A complex. PST repeat phosphorylation is dispensable for MDC1 recruitment and assembly of the MDC1–TOPBP1–CIP2A complex at mitotic DNA breaks. (A) Images of fixed, mitotic U2OS cells expressing HaloTagged MDC1 (JFX650) or knock-in of PST mutants untreated or treated with 20 µg/ml zeocin for 1 h. DNA was labeled with Hoechst, and cells were immunostained with antibodies targeting TOPBP1 and CIP2A. Summary data of the quantification of the number of (B) CIP2A, (C) TOPBP1, and (D) MDC1 foci of the experiment shown in panel (A) with three biological replicates (>50 cells per condition per replicate, red bar indicates mean, two-way ANOVA). Enrichment of MDC1 on mitotic chromatin leads to chromosome missegregation We have identified a molecular switch that displaces MDC1 from mitotic chromatin via CDK1-dependent phosphorylation of the PST repeat region. Because the PST repeat region is able to form multivalent interactions with chromatin, we hypothesized that failure to displace MDC1 from mitotic chromosomes could promote aberrant interchromosomal attachments, leading to chromosome missegregation. To test this hypothesis, we performed time-lapse imaging of mitotic progression in MDC1 knockout cells stably expressing different HaloTagged MDC1 variants. In cells expressing WT MDC1 or MDC1 ΔPST, the median time from nuclear envelope breakdown (NEB) to anaphase onset was ∼35 min (Fig. 6A and B; Movies S23 and S24). In contrast, cells expressing non-phosphorylatable (T > A) or phospho-mimicking (T > D) PST mutants exhibited prolonged mitotic duration, consistent with delays in chromosome alignment sufficient to satisfy the spindle assembly checkpoint (Fig. 6A and B; Movies S25–S29). Notably, accumulation of MDC1 on mitotic chromatin varied depending on both PST phosphorylation status and expression levels. In transient overexpression experiments (Fig. 1L), MDC1 ΔPST and WT MDC1 showed minimal chromatin association, while the T > D and T > A mutants displayed progressively higher levels of accumulation. In contrast, under the doxycycline-inducible expression conditions used for live-cell imaging (Fig. 6A), we observed modest enrichment of WT MDC1 on mitotic chromatin in addition to the T > D and T > A mutants, consistent with elevated protein levels overcoming the reduced affinity of phosphorylated MDC1 for chromatin. Accumulation of MDC1 on mitotic chromosomes leads to chromosome missegregation. (A) Still images from live-cell time-lapse imaging of U2OS ΔMDC1 stably expressing HaloTagged MDC1 variants labeled with JFX650. (B) Quantification of the time from nuclear envelope breakdown to anaphase onset of U2OS ΔMDC1 cells stably expressing HaloTagged MDC1 variants. (C) Quantification of time to anaphase onset of endogenous PST KI mutants (n = 200 cells for each group, red bar indicates median. Data compared by one-way ANOVA). (D) Quantification of the time from nuclear envelope breakdown to anaphase onset, the chromosome segregation phenotype, and Halo-MDC1 expression level of U2OS ΔMDC1 cells stably expressing HaloTagged MDC1 variants labeled with JFX650. The chromosome segregation phenotype was determined by manual inspection. To more precisely determine the effects of PST phosphorylation independent of overexpression, we performed similar time-lapse imaging of endogenous knock-in cell lines which express MDC1 variants at near-physiological levels (Figs 5A and 6C; Supplementary Fig. S3F). In this context, cells expressing WT MDC1, MDC1 knockout, ΔPST, and T > D variants all progressed through mitosis with similar timing, while the T > A mutant exhibited a significant delay in anaphase onset (Fig. 6C). Notably, chromosome missegregation events were rare across all genotypes at endogenous expression levels. In contrast, inducible overexpression of MDC1 led to a marked increase in chromosome missegregation events (Fig. 6A and Movies S23–S29), including anaphase bridges and complete failure to separate chromosome masses during anaphase onset, which is consistent with aberrant chromosome attachments. The highest frequency of missegregation events was observed in cells expressing the T > A mutant (∼46%), followed by T > D (∼20%) and WT MDC1 (∼13%) (Fig. 6D). Consistent with a chromatin-binding-dependent mechanism, missegregation events were rarely observed in cells expressing MDC1 ΔPST, which is excluded from mitotic chromosomes (Fig. 6A and D). Moreover, the frequency of chromosome missegregation positively correlated with MDC1 expression levels (Fig. 6D), indicating that these defects are driven by increased chromatin engagement and exacerbated by loss of PST phosphorylation. Together, these observations demonstrate that accumulation of MDC1 on mitotic chromatin promotes aberrant chromosome interactions that impair proper segregation, particularly when MDC1 levels are elevated. These findings support a model in which phosphorylation-dependent displacement of MDC1 from mitotic chromosomes is required to limit multivalent chromatin interactions mediated by the PST repeat region and ensure faithful chromosome segregation. The PST repeat region of MDC1 is sufficient for chromatin tethering Our results demonstrate that accumulation of MDC1 on mitotic chromosomes can lead to chromatin attachments that interfere with proper chromosome segregation (Fig. 6). This phenotype was most pronounced in cells expressing the non-phosphorylatable (T > A) variant of MDC1, which is expected to have the highest affinity for chromatin in mitotic cells. This suggests that the ability of MDC1 to cause chromosome missegregation scales with its affinity for chromatin. To test whether local enrichment of the non-phosphorylatable (T > A) PST repeat region of MDC1 is sufficient to drive aberrant non-centromeric chromosome attachments and chromosome missegregation, we fused the WT and T > A PST repeat regions to the Lac inhibitor (LacI) and GFP and expressed these constructs from a doxycycline-inducible expression cassette in U2OS cells carrying a LacO array on chromosome 1 containing >50,000 LacI binding sites [40]. This system enabled targeted enrichment and simultaneous visualization of the PST repeat constructs at a single chromosomal locus (Fig. 7A). Following induction of LacI-PST-GFP fusion expression, we performed time-lapse imaging of mitotic cells to determine whether the PST repeat region was sufficient to trigger missegregation of the chromosome carrying the LacO array (Fig. 7A). In cells expressing the WT PST repeat region fused to LacI, sister chromatids carrying the LacO array were equally segregated to the daughter cells, with a single signal being detected within each chromosome mass upon anaphase onset (Fig. 7B and C; Movie S30). In contrast, expression of the T > A PST repeat fusion led to frequent segregation defects, including lagging or fragmented LacO arrays during anaphase, unequal segregation of the LacO locus, and persistence of the LacO signal within the cleavage furrow, often resulting in cytokinesis failure (Fig. 7B and C; Movies S31–S34). These findings demonstrate that local enrichment of the non-phosphorylatable PST repeat region is sufficient to promote aberrant sister chromatid attachments and drive chromosome missegregation, independent of the full-length MDC1 protein. Local accumulation of the non-phosphorylatable T > A PST repeat region is sufficient to trigger chromosome missegregation. (A) Experimental design for local enrichment of the PST repeat region via the LacI–LacO interaction. (B) Still images from time-lapse imaging of U2OS cells containing a single LacO array expressing WT and T > A LacI-PST repeat-GFP fusion proteins. (C) Quantification of the chromosome segregation phenotype of U2OS cells containing a single LacO array expressing WT and T > A LacI-PST repeat-GFP fusion proteins. Errors include lagging LacO arrays, LacO missegregation, and failed cytokinesis. The PST repeat region is required for homologous recombination Previous studies, including our own, have demonstrated that the PST repeat region of MDC1 contributes to DSB repair through HR, although the underlying mechanism remains unclear. To determine the contribution of the PST repeat region to DNA DSB repair, we performed clonogenic survival assays treating with zeocin using cell lines expressing MDC1 variants from its endogenous locus (Supplementary Fig. S3F). Cells expressing MDC1 ΔPST exhibited increased sensitivity to zeocin compared to WT controls (Fig. 8A), indicating a defect in the response to zeocin-induced DSBs. Cells expressing the T > A and T > D mutants showed only modest increases in sensitivity relative to WT (Fig. 8A). Notably, we were unable to detect PST phosphorylation in interphase cells (Fig. 3C–E), consistent with the limited impact of T > A and T > D mutations under these conditions. To determine whether the sensitivity of MDC1 ΔPST cells reflects impaired HR, we performed clonogenic survival assays using the PARP inhibitor olaparib. Consistent with a defect in HR, MDC1 ΔPST cells were also sensitive to olaparib (Supplementary Fig. S4A). The PST repeat region is required for DNA break repair by HR. (A) Quantification of clonogenic survival assays using zeocin of U2OS cells expressing HaloTagged MDC1 and endogenous knock-in of PST mutants. (N = 3 biological replicates plated in triplicate, mean ± S.D., two-way ANOVA). (B) Images of fixed interphase U2OS cells expressing HaloTagged MDC1 or MDC ΔPST from its endogenous locus untreated or treated with zeocin to induce DNA breaks. Cells were stained with JFX650 HaloTag-ligand and Hoechst and immunolabeled with antibodies targeting RAD51 and γH2AX. Cells were treated with 20 µg/ml zeocin for 4 h prior to fixation. (C) Quantification of the number of RAD51 foci of the experiment shown in panel (B) with three biological replicates (>100 cells per condition, median ± S.D., two-way ANOVA). (D) Images of fixed interphase U2OS cells expressing Halo-MDC1 from its endogenous locus, MDC1 knock-out cells (ΔMDC1), and U2OS ΔMDC1 stably expressing HaloTagged MDC1 variants untreated or treated with 20 µg/ml zeocin for 4 h to induce DNA breaks. Cells were stained with JFX650 HaloTag-ligand and Hoechst and immunolabeled with an antibody targeting RAD51. (E) Quantification of the number of RAD51 foci of the experiment in panel (C). An individual replicate is plotted on the left (red line = median, >100 cells per condition) and three biological replicates are plotted on the right (median ± S.D., two-way ANOVA). (F) Images of fixed interphase U2OS cells expressing Halo-MDC1 from its endogenous locus, MDC1 knock-out cells (ΔMDC1), and U2OS ΔMDC1 cells stably expressing HaloTagged MDC1 variants untreated or treated with 20 µg/ml zeocin for 4 h to induce DNA breaks. Cells were stained with JFX650 HaloTag-ligand and Hoechst and immunolabeled with an antibody targeting 53BP1. (G) Quantification of the number of 53BP1 foci of the experiment in panel (E). An individual replicate is plotted on the left (red line = median, >100 cells per condition) and three biological replicates are plotted on the right (median ± S.D., two-way ANOVA). Quantification of the accumulation of (H) Halo-MDC1 and Halo-MDC1 ΔPST and (I) ubiquitin Reader1.0 at LMI-induced DNA lesions in U2OS ΔMDC1 cells stably expressing indicated HaloTagged MDC1 variants (N = 10–19 cells, mean ± 95% CI). (J) Quantification of the maximal accumulation of ubiquitin Reader1.0 at LMI-induced DNA lesions in U2OS ΔMDC1 cells stably expressing indicated HaloTagged MDC1 variants (N = 10–19 cells, mean ± 95% CI, one-way ANOVA). (K) Quantification of HR efficiency using the DR-GFP reporter stably integrated into the AAVS1 locus of U2OS ΔMDC1 cells. I-Sce and MDC1 variants were transiently transfected (N = 3–4 biological replicates, mean ± S.D., two-way ANOVA). To further define the role of the PST repeat region in HR, we analyzed MDC1 and RAD51 foci formation in cells expressing MDC1 variants from the endogenous locus (Fig. 8B and Supplementary Fig. S4B and C). WT, T > A, T > D, and ΔPST MDC1 formed robust foci in response to zeocin treatment, indicating that these mutations do not impair MDC1 recruitment to sites of damage in interphase (Fig. 8B and Supplementary Fig. S4B). In contrast, although MDC1 ΔPST retained the ability to localize to DNA breaks, it resulted in a pronounced defect in RAD51 foci formation (Fig. 8B and C; Supplementary Fig. S4C). Conversely, cells expressing T > A or T > D PST mutants displayed RAD51 foci formation comparable to the WT PST knock-in control (Supplementary Fig. S4C). Treatment with zeocin for 4 h did not change the cell cycle profiles of the cell lines, ruling out the possibility that alterations in cell cycle progression underlie these observations (Supplementary Fig. S4D and E). These data indicate that the PST repeat region functions to support RAD51 loading but that its phosphorylation state does not impact this function in interphase. To further confirm the contribution of the PST repeat region to RAD51 recruitment, we next examined RAD51 foci formation in ΔMDC1 cells with inducible expression of MDC1 variants. Under these conditions, WT MDC1 robustly supported RAD51 foci formation, whereas deletion of either the PST repeat region or the BRCT domain abolished RAD51 foci formation (Fig. 8D and E). In contrast, 53BP1 foci formation was largely unaffected by deletion of the PST repeat region but was abolished by deletion of the BRCT domain (Fig. 8F and G), indicating that PST-dependent functions are specifically required for HR but are dispensable for 53BP1-dependent end-joining. Ubiquitin signaling mediated by RNF8 and RNF168 downstream of MDC1 is critical for the recruitment of HR (e.g. BRCA1) and NHEJ (e.g. 53BP1) effectors [16, 61]. To analyze the role of the PST repeat region and BRCT domains of MDC1 in ubiquitin signaling, we monitored chromatin ubiquitination using a genetically encoded H2AK13/15-ubiquitin sensor (Reader1.0-eGFP) fused to a nuclear localization sequence (NLS) to detect RN168 activity in living cells [44]. The NLS-Reader1.0-eGFP fusion protein was stably expressed in ΔMDC1 knock-out cells and cell lines with inducible expression of HaloTagged WT MDC1, MDC1 ΔPST, and MDC1 ΔBRCT. To simultaneously analyze Halo-MDC1 recruitment and downstream ubiquitin signaling, we performed LMI experiments to induce localized DNA damage. Consistent with ubiquitination occurring downstream of MDC1 recruitment, Reader 1.0 accumulation was delayed compared to MDC1 enrichment (Fig. 8H and I, Supplementary Fig. S4F and G, and Movies S35–S38). A low level of chromatin ubiquitination was detected in MDC1 knockout cells (Fig. 8I, Supplementary Fig. S4H, and Movie S38), suggesting the presence of a limited MDC1-independent ubiquitination mechanism. Compared to WT MDC1, H2AK13/K15 ubiquitination was reduced to ∼70% and ∼44% in cells expressing MDC1 ΔPST or MDC1 ΔBRCT, respectively (Fig. 8J and Supplementary Fig. S4H). Thus, although the PST repeat region contributes to efficient ubiquitin signaling, substantial ubiquitination is retained in its absence. The inability of MDC1 ΔPST to support RAD51 loading could therefore reflect reduced ubiquitin signaling. However, because the residual ubiquitination in ΔPST cells is sufficient to support 53BP1 recruitment, these results suggest that RAD51 loading either requires a higher threshold of ubiquitination or depends on an additional activity of the PST repeat region. Given our findings that the PST repeat region mediates multivalent chromatin interactions, we hypothesized that it may facilitate HR by promoting DNA tethering or stabilizing interactions with the sister chromatid during repair. To directly test whether altering PST-mediated chromatin binding affects HR efficiency, we performed DR-GFP reporter assays in MDC1 knockout cells harboring knock-in of the reporter in the AAVS1 locus and transiently complemented with MDC1 variants. In previous work we demonstrated that MDC1 ΔPST and MDC1 ΔBRCT exhibited similar defects in HR using this approach (Fig. 8K) [38]. The T > A and T > D MDC1 variants also displayed a significant defect in HR compared to WT MDC1 but supported higher levels compared to the ΔPST and ΔBRCT mutants (Fig. 8K). These observations suggest that altering the chromatin binding properties of the PST repeat region reduces the ability of MDC1 to facilitate HR. Discussion The results presented here identify a critical function for the PST repeat region of MDC1 in genome stability. We demonstrate that the PST repeats form a multivalent chromatin-binding module that is sufficient to tether chromatin. In interphase cells, this activity supports DSB repair by HR, acting downstream of MDC1 recruitment to promote efficient RAD51 loading. In contrast, in mitosis this chromatin-binding activity must be attenuated by phosphorylation, and failure to regulate this interaction leads to delayed mitosis, aberrant chromosome attachments, and missegregation. Together, these findings establish the PST repeat region as a tunable chromatin interaction module that supports DNA repair in interphase while requiring tight regulation during mitosis to support proper chromosome segregation. The PST repeat region is a multivalent chromatin interaction domain with tunable affinity Observations by Salguero et al. demonstrated the PST repeat region of MDC1 interacts with the acidic patch of nucleosomes [39]. Consistent with this model, our previous work using live-cell single-molecule imaging demonstrated that MDC1 is constitutively anchored to chromatin by the PST repeat region [38]. In this work, we found that reducing the number of the PST repeats progressively decreases chromatin association, suggesting that individual repeats contribute additively to binding and that the PST repeat region engages multiple nucleosomes simultaneously. These observations are consistent with the idea that the PST repeat region contributes to local chromatin architecture through multivalent interactions with nucleosomes. We further show that PST-chromatin interactions are regulated during mitosis by CDK1-Cyclin B-dependent phosphorylation of conserved TP motifs within the repeats. Overexpression of WT, T > D, and T > A MDC1 can result in chromatin association of MDC1 in mitotic cells, while deletion of the PST repeats results in complete exclusion of MDC1 from mitotic chromosomes. Together, these results indicate that phosphorylation attenuates but does not completely eliminate PST-mediated chromatin binding. We conclude that PST repeat-mediated chromatin engagement is governed by a balance between phosphorylation state and protein dosage. While we did not observe phosphorylation of the TP sites within the PST repeats in interphase cells, it is interesting to speculate that other posttranslational modifications of the repeats and chromatin context fine-tune the PST repeat-chromatin interaction in certain situations (e.g. DNA replication) or predispose DSBs to repair by a specific pathway. Importantly, we also address the functional relevance of mitotic phosphorylation of the PST repeat region of MDC1. Our experiments demonstrate that the T > A PST mutant is strongly enriched on mitotic chromosomes and, in the context of overexpression, leads to a highly penetrant chromosome missegregation phenotype characterized by frequent failure to separate chromosome masses during anaphase. This phenotype is correlated with the expression level of MDC1 and is observed to a lesser degree in cells expressing WT and T > D PST alleles. Notably, analysis of endogenous knock-in cell lines revealed that the T > A mutant also exhibits a significant delay in the time to anaphase onset, despite the absence of overt chromosome missegregation events under these conditions. These findings indicate that disruption of PST phosphorylation is sufficient to impair mitotic progression at physiological expression levels, while the severity of chromosome segregation defects can be further amplified by increased MDC1 abundance. The observation that targeted enrichment of the T > A PST repeat region alone is sufficient to induce chromosome missegregation indicates that this phenotype arises from the intrinsic properties of the repeat sequence, independent of other domains of MDC1. These results further suggest that MDC1 expression levels must be finely tuned to balance its role in DNA repair with the need to prevent aberrant chromatin interactions during mitosis. MDC1 in mitotic DNA break tethering In mitotic cells, MDC1 has been shown to contribute to tethering of DSBs to maintain their connectivity throughout cell division, allowing them to be repaired in the following G1-phase by 53BP1 dependent end-joining [30]. This function is mediated in part through recruitment of TOPBP1 via a direct interaction near the N-terminus of MDC1 [30]. TOPBP1 in turn recruits CIP2A, which is excluded from nuclei in interphase cells, to DNA breaks in mitosis [31, 62, 63]. The MDC1–TOPBP1–CIP2A complex has also been linked to the clustering of shattered chromosome fragments derived from micronuclei, although the precise contribution of MDC1 to this process is still not entirely clear due to conflicting observations [64, 65]. The molecular mechanism by which this complex maintains DNA break connectivity is not fully understood. Recruitment of MDC1 to sites of damage via its BRCT domain provides a foundation for assembly of the complex, but how chromatin interactions are stabilized at these sites remains unclear. Our results demonstrate that the PST repeat region is required for efficient accumulation of MDC1, TOPBP1, and CIP2A at mitotic DNA breaks, suggesting that PST-mediated chromatin interactions contribute to stable retention of MDC1 at these sites. In addition, our findings indicate that PST-dependent chromatin binding can influence chromatin behavior in mitosis in a dose- and affinity-dependent manner. Although phosphorylation of the PST repeat region reduces chromatin association under physiological conditions, increased local accumulation of MDC1 may be sufficient to overcome this regulation and promote chromatin engagement. These observations raise the possibility that PST-mediated chromatin interactions contribute to altering local chromatin architecture at mitotic DNA breaks, rather than acting as a direct physical tether between DNA ends. Future experiments uncoupling break recruitment from multivalent chromatin binding by the PST repeat region will be necessary to directly test this possibility. Our observations demonstrate that even phosphorylated WT MDC1 can be tethered to mitotic chromatin when overexpressed. Therefore, it is a reasonable possibility that the PST repeats of MDC1 could play an important role in chromatin tethering in the context of mitotic MDC1–TOPBP1–CIP2A bound DNA breaks. If the PST repeat region of MDC1 physically tethers together DNA, CIP2A, which is a well-established regulator of protein phosphorylation via its interaction with protein phosphatase 2A and B56α [66], could be required to tune the affinity of the PST repeat region for chromatin by regulating its phosphorylation status to ensure effective DNA break tethering. CIP2A is typically thought of as an inhibitor of phosphatase activity and the precise role of CIP2A in regulating the phosphorylation of MDC1’s PST repeat region must be clarified in future work. Tethering of MDC1 to chromatin contributes to DNA break repair by homologous recombination Previous work by others, together with our results, demonstrates that the PST repeat region of MDC1 contributes to DSB repair by HR [10, 38, 67]. However, the mechanism by which the PST repeats facilitate HR remains unclear. MDC1 contributes to both HR and NHEJ by initiating a ubiquitination cascade that results in the recruitment of BRCA1 and 53BP1, respectively [8]. BRCA1 facilitates the recruitment of PALB2 and BRCA2 to DSBs, which loads RAD51 onto the exposed single-stranded DNA end [68–70]. Given that loss of the PST repeat region selectively impairs RAD51 loading while preserving 53BP1 recruitment, our findings raise the possibility that PST-mediated chromatin interactions help bias repair pathway choice by promoting a chromatin environment that favors HR over end joining. Our results demonstrate that ubiquitin signaling is reduced in cells expressing MDC1 lacking the PST repeat region. Notably, this reduced signaling remains sufficient to support 53BP1 recruitment but fails to support efficient RAD51 loading. In addition, previous work by others has demonstrated that BRCA1 recruitment to DSBs is unaffected by deletion of the PST repeat region [10]. This suggests that a molecular event downstream of BRCA1 recruitment but prior to RAD51 loading requires the PST repeat region of MDC1. One possibility is that reduced ubiquitin signaling in the absence of the PST repeat region is insufficient to support efficient RAD51 loading. However, given that 53BP1 recruitment is largely preserved, these observations also raise the possibility that an additional function of the PST repeat region, such as its ability to mediate multivalent chromatin interactions, contributes to this step in HR. We propose that the PST repeat region of MDC1 promotes chromatin interactions in the vicinity of DSBs that support efficient HR. While the MRN complex has well-established DNA end-tethering activity during DSB repair, our findings suggest that MDC1, through its PST repeat-mediated multivalent chromatin interactions, may act in parallel to modulate the local chromatin architecture that stabilizes break-associated assemblies, raising the possibility that these activities functionally complement each other [71]. Consistent with this model, HR efficiency is strongly reduced in the absence of the PST repeat region, while cells expressing phospho-mimicking or non-phosphorylatable variants of the PST repeats exhibit more modest reductions. These findings suggest that altering the affinity of the PST repeats for chromatin impairs HR, and that this interaction may be dynamically regulated in some way during the repair process. One possibility is that, in the absence of the PST repeat region, chromatin surrounding DNA breaks is not maintained in a configuration that supports efficient RAD51 loading, leading to reduced formation of RAD51 foci. Importantly, both T > A and T > D variants of MDC1 have very modest reductions in HR efficiency, indicating that proper regulation of PST-mediated chromatin binding, rather than simply its presence or absence, may be critical for this process. These observations raise the possibility that the PST repeat region plays an architectural role through multivalent chromatin interactions, either by enhancing the efficiency of DNA damage-induced ubiquitin signaling or by promoting a chromatin environment that is permissive for RAD51 loading. In summary, our findings establish the PST repeat region of MDC1 as a multivalent chromatin interaction module whose activity is differentially regulated across the cell cycle. This tunable chromatin-binding activity contributes to efficient HR in interphase and must be attenuated during mitosis to prevent aberrant chromatin interactions that compromise chromosome segregation. More broadly, these results highlight how regulated chromatin engagement by DNA damage response factors can coordinate repair with cell cycle progression to preserve genome stability. Supplementary Material Acknowledgements We thank the Luke Lavis lab (HHMI Janelia Research Campus) for providing Janelia Fluor ligands. All flow cytometry data were obtained using instrumentation in the MSU Flow Cytometry Core Facility, and we especially thank Matthew Bernard and Daniel Vocelle for providing service regarding cell sorting and flow cytometry analysis. We thank the MSU IQ Microscopy Core Facility for supporting our work. Author contributions: Conceptualization: J.R.H and J.C.S.; Experiments: J.R.H, M.M., C.P., S.M., and G.I.P.; Data Analysis: J.R.H., D.G.B., C.P., S.M., and J.C.S.; Writing: Original Draft: J.R.H. and J.C.S.; Writing: Review and Editing: J.R.H. and J.C.S. Contributor Information Joshua R Heyza, Institute for Quantitative Health Science and Engineering, Michigan State University, East Lansing,MI 48824, United States; Department of Pharmacology, Wayne State University School of Medicine, Detroit, MI 48202, United States; Institute of Environmental Health Sciences, Wayne State University, Detroit, MI 48202, United States. Mariia Mikhova, Institute for Quantitative Health Science and Engineering, Michigan State University, East Lansing,MI 48824, United States; Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI 48824, United States. Cody Phillips, Institute of Environmental Health Sciences, Wayne State University, Detroit, MI 48202, United States. Shomaila Mehmood, Institute of Environmental Health Sciences, Wayne State University, Detroit, MI 48202, United States. Gloria I Perez, Institute for Quantitative Health Science and Engineering, Michigan State University, East Lansing,MI 48824, United States. David G Broadbent, Institute for Quantitative Health Science and Engineering, Michigan State University, East Lansing,MI 48824, United States; College of Osteopathic Medicine, Michigan State University, East Lansing, MI 48824, United States. Jens C Schmidt, Institute for Quantitative Health Science and Engineering, Michigan State University, East Lansing,MI 48824, United States; Department of Obstetrics, Gynecology and Reproductive Biology, Michigan State University, East Lansing, MI 48824, United States. Supplementary data Supplementary data is available at NAR online. Conflict of interest None declared. Funding This work was supported by grants from the National Institutes of Health (F32GM139292, R00GM149942) to J.R.H. and (DP2 GM142307) to J.C.S. The flow cytometry facility is funded in part through the financial support of Michigan State University’s Office of Research & Innovation and Colleges of Osteopathic Medicine, Human Medicine, Veterinary Medicine, Natural Sciences, and Engineering. The Wayne State University Microscopy, Imaging, and Cytometry Resources Core is supported in part by National Institutes of Health Center grant P30CA22453 to the Karmanos Cancer Institute and Wayne State University. Funding to pay the Open Access publication charges for this article was provided by grants from the National Institutes of Health (F32GM139292, K99GM149942, R00GM149942) to J.R.H. and (DP2 GM142307) to J.C.S. Data availability Raw, uncropped images of gels and blots presented in this paper are included in the supplemental data file. All primary data will be made available upon reasonable request by J.C.S. (schmi706@msu.edu). References Associated Data Supplementary Materials Data Availability Statement Raw, uncropped images of gels and blots presented in this paper are included in the supplemental data file. All primary data will be made available upon reasonable request by J.C.S. (schmi706@msu.edu).