Staining of human cells with FAM111B antibodies revealed that the protease can be found in both cytoplasm and nucleus; however, the majority of FAM111B resided in the latter compartment (but not in nucleoli) (Figure 1A). The modulation of FAM111B protein levels, dependent on cell cycle position (enriched in S-phase cells with characteristic PCNA replication foci) was very apparent and consistent with previously published reports (Sun et al., 2019; Hoffmann et al., 2020; Kawasaki et al., 2020) (Figure 1A). Fractionation of U2OS cells showed that FAM111B was largely soluble, with a fraction of the protease also bound to insoluble components of the cell such as chromatin or cellular membranes, particularly after blocking the cell cycle in early S-phase with hydroxyurea (Figure 1B). To determine whether the immobilized fraction of the FAM111B accumulated in a particular nuclear compartment, we analyzed FAM111B staining in U2OS cells using high resolution microscopy. Prior to analysis, soluble and proteins weakly bound to the structural components of the cell (such as cytoskeleton, membranes or chromatin, etc.) were removed by detergent extraction and the remaining cellular material stained with antibodies against FAM111B and Lamin A/C. FAM111B staining in these conditions revealed a clear three-dimensional organization inside the nucleus (Figure 1C and Suppelementary Movie S1A), again no nucleolar staining was observed. Interestingly, FAM111B also localized to the nuclear periphery where it either interacted with or was found embedded in the nuclear lamina (Figure 1C, white arrow heads, Suppelementary Movies S1A, S1B). This suggested that FAM111B interacted with the lamin network. To test whether FAM111B directly interacted with lamin proteins, we characterized FAM111B binding partners using mass spectrometry. Analysis of the FAM111B interactome revealed that FAM111B did not form complexes with lamins but instead with components of the nuclear pore complexes (NPCs) (Figure 1D and Supplementary Figure S1A). Co-immunoprecipitation experiments with FLAG-FAM111B confirmed that the protease interacted with two nucleoporins SEC13 and NUP42 (Figure 1E). This was consistent with the observed localization of FAM111B and suggested that FAM111B may be recruited to the nuclear periphery through the interaction with components of the NPCs.

Localization of the endogenous FAM111B protease at the nuclear periphery and its interactions with the nuclear pore complexes suggested that this localization may be important for FAM111B function. We wondered whether different variants of FAM111B including mutation found in the HFP syndrome could affect this localization. We over-expressed different FAM111B variants as FLAG-fusions in U2OS cells, extracted the soluble proteins with detergent and stained the cells with FAM111B antibodies (Figure 2A). We have observed three distinct staining patterns which we quantified as: pan nuclear, peripheral enriched and exclusively peripheral (Figure 2A). FLAG-tagged FAM111B wild-type (WT) showed similar staining pattern to the endogenous FAM111B (Figure 2A, pan nuclear and B), whereas in the cells over-expressing the HFP variant (Q430P) we could observe a small fraction of the cells where FLAG-FAM111B markedly accumulated at the nuclear periphery (Figure 2A, peripheral enriched and B). Because the HFP mutant variants of FAM111B undergo auto-cleavage into proteolytic fragments (Figures 1E, 2D), which might be relevant to their dominant negative function and disease development (Hoffmann et al., 2020), we investigated whether the resulting fragments had similar potential to localize to the nuclear periphery. The auto-cleavage site in the FAM111A protease (F334) is conserved in FAM111B (F438, Figure 2C) (Kojima et al., 2020), and the theoretical molecular weight of the N-terminal fragment after cleavage at this site corresponded to the actual mass detected by immunoblotting (Figures 1E, 2D). We cloned both WT predicted fragments: N-terminal cleavage fragment (NCF) and C-terminal cleavage fragment (CCF) and over-expressed them in U2OS cells as FLAG-fusion proteins. The NCF fragment showed a similar localization to the HFP mutation with partial accumulation at the periphery (Figure 2A, peripheral enriched and B) whereas the C-terminal fragment did not express to detectable levels. Cellular fractionation of the cells over-expressing FAM111B with the disease relevant Q430P (QP) mutation showed that the N-terminal auto-cleavage fragment accumulated exclusively in the insoluble fraction of the cell (Figure 2D). Furthermore, we observed that FAM111B lacking a centrally located domain which is not conserved between FAM111A and FAM111B (FAM111B extra domain, Δ279-386aa, ED) demonstrated exclusive accumulation of the FAM111B protease at the nuclear periphery (Figure 2A, exclusively peripheral, and B). Our analysis suggested that FAM111B activity might be transiently required at the nuclear periphery and that accumulation of HFP associated Q430P mutation or truncated forms of FAM111B at this location could be potentially pathological.

Because FAM111B localized to the nuclear lamina we predicted that loss of FAM111B may have a negative effect on nuclear morphology. To test this hypothesis, we employed a reverse genetic approach and manipulated the FAM111B gene in human cells (U2OS and MCF7) using Cas9-mediated editing and two independent guide RNAs (Figure 2E, top panel). Because CRISPR-Cas9 engineering can generate chromosomal rearrangements (Rayner et al., 2019), we screened FAM111B locus integrity in the targeted clones using fluorescent in situ hybridization (FISH) probes, and selected those which had a karyotype identical to wild-type cells and the intact probe pattern at the targeted FAM111B loci [Supplementary Figure S1B, analogous data for MCF7 clones (not shown)]. As previously reported (Blomen et al., 2015; Hart et al., 2015; Wang et al., 2015; Hoffmann et al., 2020), loss of FAM111B expression was compatible with cellular viability; however, U2OS cells lacking FAM111B showed approximately 2-fold growth defect characterized by accumulation of cells in G1 phase of the cell cycle (Figure 2E bottom panel and F). Staining of FAM111B ^−/- cells with antibodies against lamins showed that the FAM111B-deficient cells had increased levels of nuclear abnormalities such as abnormal shape (including nuclear blebbing) and doughnut nuclei (Figures 3A, B), suggesting that FAM111B function at the nuclear lamina may be required for the maintenance of normal nuclear shape. Similar phenotypes are observed in cells with lamin mutations or other genetic backgrounds caused by deficiencies in the components of the lamina network [reviewed in (Dobrzynska et al., 2016)]. We tested whether loss of FAM111B protease expression had any effect on the stability of lamin proteins. We found normal levels of Lamin B1 and only a small change in the Lamin A to C ratio in U2OS cells lacking FAM111B (Figure 3C), suggesting that the protease is unlikely involved in direct processing or regulation of lamin stability.

The association of FAM111B with nuclear pore complexes suggested a potential role of the protease in the nuclear transport. We sequenced mRNAs isolated from nuclear and cytoplasmic fractions of wild-type and FAM111B KO cells and found no major differences in the cellular mRNA content in the absence of FAM111B (data not shown). This suggested that global mRNA transport is not affected by the loss of FAM111B expression; however, we cannot rule out the possibility that transport of specific mRNAs or proteins may be affected in the absence of FAM111B. Besides NPCs canonical function as pores, they also serve as hubs for recruitment of chromatin and nuclear factors required for RNA synthesis, DNA repair and telomere maintenance [reviewed in (Lamm et al., 2021)]. Nuclear pore complexes participate in maintenance of genome stability through the recruitment of DNA repair intermediates such as double strand breaks and collapsed replication forks (Khadaroo et al., 2009; Whalen et al., 2020). Therefore, we wondered whether similarly to nuclear pore complex mutants (Palancade et al., 2007; Gaillard et al., 2019; Kramarz et al., 2020), cells lacking FAM111B could be DNA repair deficient. We treated wild-type and mutant cells with compounds that induce DNA damage and with compounds that interfere with cell cycle, DNA replication or transcription. We analyzed survival of wild-type and mutant cells using a resazurin viability assay. Because cell survival experiments rely on proliferation and measurement of relative metabolic activity of the cells and U2OS FAM111B ^−/- mutants have a growth defect compared to WT, we focused only on MCF7 FAM111B-null cells, which we found to be no more sensitive than their wild-type counterparts to multiple drugs tested (Supplementary Figures S2A–J). This indicated that FAM111B was not directly involved in DNA repair, replication or transcription in MCF7 cells. Even though FAM111B ^−/- cells were proficient in their responses to DNA damage, we wondered whether their abnormal nuclear shape was associated with increased levels of endogenous DNA damage as is observed in mutants of the components of the nuclear lamina (Gonzalez-Suarez et al., 2011; Kreienkamp et al., 2016; Earle et al., 2020). We thus screened the mutant cells for hallmarks of genomic instability. Quantification of micronuclei in the mutant cells (Figure 3D) showed greater levels of these events in the absence of FAM111B protease compared to wild-type cells. Consistently, we could also observe higher frequencies of events upstream of micronuclei formation such as increased mitotic DNA synthesis (MIDAS) foci (Figures 3E, F) and greater levels of BLM-coated DNA ultra-fine bridges (UFBs) in U2OS cells (Figures 3G, H). Re-introduction of wild-type but not protease dead (PD) variant into the FAM111B ^−/- cells could rescue the normal levels of UFBs (Figure 3I and Supplementary Figure S2L). This confirmed that the observed phenotype is specific to the loss of FAM111B and that protease activity of FAM111B plays an essential role in maintaining low levels of BLM bridges in U2OS cells. In the same assay, the disease Q430P variant did not restore WT levels, suggesting that HFP Q430P mutation is at least partially defective in this function (Figure 3I and Supplementary Figure S2L). We conclude that FAM111B-deficient cells, even though proficient in DNA repair accumulate DNA damage and show hallmarks of genomic instability.

Beside the role that the nuclear pore complexes play in the responses to DNA damage they are also involved in the telomere lengthening and repair by recruiting telomeres to the nuclear envelope (Khadaroo et al., 2009). Interestingly, HFP patients develop pulmonary fibrosis which strongly correlates with short telomeres [reviewed in (Goussot et al., 2017; Arowolo et al., 2022; Hoeger, 2022; Macchiaiolo et al., 2022; Wu et al., 2022) and (Armanios, 2022)]. To test whether FAM111B may be required for telomere homeostasis, we looked at the length and condition of the telomeres in FAM111B ^−/- cells. Quantification of the telomeric FISH probe intensity in the metaphase spreads prepared from wild-type and mutant cells showed that cells lacking FAM111B protease indeed had greatly reduced telomeric DNA content (Figures 4A, C), and a similar phenotype in interphase cells was also observed (Figures 4B, C). Consistently, shorter telomeres in FAM111B ^−/- U2OS cells recruited less of the shelterin component TRF2 compared to wild-type counterparts (Figures 4D, E). These phenotypes were independent of passage number (Supplementary Figure S1D) and were not caused by off target effect of the CRISPR deletion method applied (CRISPR-Cas9 treated clones with wild-type like FAM111B levels have normal levels of TRF2 foci intensities) (Supplementary Figure S1C). Because pulmonary fibrosis observed in the FAM111B mutation carriers can be directly linked to short telomeres, we tested whether over-expression of HFP variant Q430P in U2OS cells had any effect on telomeres. Telomere shortening requires multiple rounds of division to be detectable and unfortunately, prolonged over-expression of FAM111B (wild-type or mutant but not protease dead (Hoffmann et al., 2020), and our unpublished data) is toxic, preventing direct measurement of telomere length in extended experimental conditions. Therefore, we transiently over-expressed different variants of FAM111B in U2OS cells for no longer than 24 h and instead quantified TRF2 recruitment to the telomeres. Interestingly, overexpression of either the HFP mutant, or the wild-type protein but not the protease dead variant of FAM111B induced a rapid loss of TRF2 intensity (Figure 4F). The fact that this is detected in conditions when the telomere shortening cannot yet be observed suggests that TRF2 loss may drive telomere shortening in HFP. The fact that we see the same TRF2 reduction upon overexpression of the WT protein also implies that the normal cell cycle regulation of FAM111B, as well as the function impaired by the HFP mutation, is likely necessary for proper telomere function. Because only critically short telomeres are detrimental for cell viability and genome integrity (Hemann et al., 2001), we looked whether FAM111B ^−/- cells had any telomeric abnormalities beside short length. The same metaphase spread samples were used to determine levels of four phenotypes (Figures 5A–D, indicated by yellow arrowheads): loss of telomere signal, telomere fusion/intrachromosomal telomere signal, telomere heterogeneity and fragility. Both cell lines in the absence of FAM111B protease showed increased levels of telomere loss and fusion/intrachromosomal telomere signal events (Figures 5A, B, E), whereas telomeric heterogeneity and fragility was specifically increased only in U2OS cells (MCF7 on average showed slight decrease, Figures 5C–E). Loss of FAM111B induced telomere shortening both in telomerase positive MCF7 cells and in U2OS cells which utilize alternative lengthening of telomeres (ALT), strongly indicating that FAM111B operates independently of these telomere lengthening mechanisms. Consistent with this, levels of C-circles in the FAM111B-deficient U2OS cells were comparable to wild-type (Supplementary Figure S1E) and MCF7 knockout cells showed wild-type like survival to telomere replication inhibitor TMPyP4 (Supplementary Figure S2K).

Human osteosarcoma (U2OS), human breast cancer (MCF7) and human embryonic kidney cells (HEK293T) cells were cultured in Dulbeco’s Modified Eagle Medium (DMEM, Sigma) supplemented with 10% fetal bovine serum (Sigma), 1% glutamate (Sigma) and 1% penicillin/streptomycin (Sigma) at 37°C in humidified 5% CO₂ atmosphere. For conditional protein expression 1 μg/mL of doxycycline (Sigma) was added to the culture media and incubated for up to 24 h.

All gene cloning was performed using Gateway^® system (Invitrogen) unless stated otherwise. Briefly, human FAM111B cDNA was amplified using high fidelity Q5 DNA polymerase (New England Biolabs), cDNA template (kind gift from Dr Nicola Burgess-Brown and Dr Alejandra Fernandez-Cid) and cloned into pDONR-ZEO (Invitrogen) vector to obtain pDONR-ZEO-FAM111B-WT (wild-type). Protease dead (H490A), patient mutations (Q430P and T625N) and ΔED (Δ279-386) variants were constructed by site directed mutagenesis using pDONR-ZEO-FAM111B-WT as a template and Q5 polymerase. FAM111B cDNAs were then transferred from pDONR-ZEO to pHAGE-(N)-HA-FLAG-(PURO) vector for constitutive expression or to pSIN3A-(C)-HA-(PURO) vector for conditional expression (TET^ON) (kind gift from Prof J. Ross Chapman). FAM111B targeting constructs were cloned into pX330-(PURO) (kind gift from Prof J. Ross Chapman) using previously published protocol (Ran et al., 2013). For primer details please refer to Supplementary Material.

Primary antibodies were as follows: mouse anti-human BLM (sc-365753, Santa Cruz Biotechnology) at 1:250 (IF), mouse anti-BrdU (347580, Becton Dickinson) at 1:20 (Flow cytometry), rabbit anti-human FAM111B (HPA038637, Atlas Antibodies) at 1:1,000 (IB and IF), mouse anti-human Histone H3 (10799, Abcam) at 1:1,000 (IB), mouse anti-human Lamin A/C (sc-376248, Santa Cruz Biotechnology) at 1:1,000 (IB and IF), rabbit anti-human Lamin B1 (12987-1-AP, Proteintech Group) at 1:1,000 (IB and IF), rabbit anti-human NUP42 (16587-1-AP, Proteintech Group) at 1:1,000 (IB), mouse anti-human PCNA (sc-56, Santa Cruz Biotechnology) at 1:1,000 (IF), rabbit anti-human SEC13 (15397-1-AP, Proteintech Group) at 1:1,000 (IB), mouse anti-human TRF2 (NB100-56506, Novus Biologicals) at 1:1,000 (IF), mouse anti-human αTubulin (T6074, Sigma) at 1:10,000 (WB) or 1:1,000 (IF). Secondary antibodies were as follows: goat anti-rabbit Alexa Fluor 488 (A32731, ThermoFisher Scientific) at 1:1,000 (IF), donkey anti-mouse Alexa Fluor 555 (A32773, ThermoFisher Scientific) at 1:1,000 (IF), goat anti-rabbit HRP (31430, ThermoFisher Scientific) at 1:25,000 (IB) and donkey anti-mouse HRP (31458, ThermoFisher Scientific) at 1:25,000 (IB).

U2OS or MCF7 cells were seeded at 2 × 10⁵ onto glass cover slips in 6 well plates. 48 h post-seeding cells were fixed with 4% formaldehyde in PBS for 10 min at room temperature, washed three times with PBS (3 min each), permeabilized with CSK buffer for 5 min at room temperature (10 mM PIPES-KOH pH6.8, 100 mM NaCl, 300 mM sucrose, 1 mM EGTA, 1 mM MgCl2, 1 mM DTT and 0.5% Triton X-100) supplemented with protease and phosphatase inhibitors and again washed three times with PBS (3 min each). For detergent extraction experiments, cells were first exposed to CSK buffer for 5 min at room temperature washed three times with PBS (3 min each) and then fixed with 4% formaldehyde for 10 min at room temperature, followed by washes with PBS (3 min each). Cover slips were then blocked with 3% BSA/PBS for 30 min at room temperature, incubated with primary antibodies in the blocking buffer for 1 h at room temperature and washed three times with PBS (3 min each). This was followed by the incubation of cover slips in secondary antibodies diluted in the blocking buffer for 1 h at room temperature, three washes with PBS and mounting the cover slips onto glass slides with Vectashield + DAPI (H-1200-10, 2BeScientific). For staining of ultra-fine DNA bridges (UFBs), 2 × 10⁵ cells was seeded onto cover slips and 24 h later cells were exposed to 2 mM thymidine for 16 h. The next day, the nucleotide was washed away (three times with warm media) to release cells from the block and trapped in pro-metaphase for 3 h at 9 h post-thymidine removal with the addition of nocodazole at 200 ng/mL. Microtubule poison was then washed away (three times with warm fresh media) and cells harvested 30–45 min after release from the nocodazole block. For the staining of UFBs in the FAM111B rescue cell lines, doxycycline was added exactly 24 h prior to 30 min release mark from the nocodazole block. For detection of EdU foci in prometaphase cells (MIDAS), U2OS cells were seeded the same way as in for analysis of UFBs, however; cells were not released from the nocodazole block, labelled for last 30 min with 10 μM EdU and fixed with 4% formaldehyde for 10 min at room temperature. EdU detection was performed according to the manufacturer’s instructions (Invitrogen). Analysis was performed using wide-field Leica DM4000 or AiryScan 2 (Zeiss) microscopes.

For transient transfections of U2OS or HEK293T cells, 70%–80% confluent cultures in 6 well plates or 10 cm dishes were transfected using 3 µg or 18 µg of linear PEI (Sigma), 0.2 mL or 1 mL OPTI-MEM medium (Life Technologies) and 1 µg or 6 µg of DNA, respectively. Cells were analyzed by microscopy or harvested 24 h post-transfection. To obtain stably expressing U2OS clones, 70%–80% confluent cultures in 10 cm dishes were transfected as above and 24 h post-transfection cells were exposed to Puromycin (Invitrogen) at 0.5 μg/mL. After approximately 2-3 weeks, single colonies were collected by trypsinization and expanded.

As described in the section “Cell transfections,” pX330-(PURO)-FAM111B gRNA #1 and #2 constructs were delivered to U2OS and MCF7 by transfections with linear PEI in 10 cm dishes. After 24 h post-transfection, cells were pulsed with puromycin at 1 μg/mL for 72 h. Approximately 2-3 weeks later, single colonies were collected by trypsinization and expanded in 24 and then 6 well plates. When clones reached 90% confluency in 6 well plates, half of the cultures was frozen, and the remaining cells harvested and analyzed by immuno-blotting with antibodies against FAM111B and α−Tubulin as a loading control. Clones negative for FAM111B expression were revived from liquid nitrogen and expanded.

Approximately 2 × 10⁶ of U2OS cells (either transiently transfected or not) was used for these experiments. Cells were washed once with ice cold PBS buffer, split in half (fractionation and whole lysate sets) and spun at 300 g for 3 min. Fractionation set was re-suspended in 100 μL of buffer A (10 mM HEPES pH7.4, 10 mM KCl, 1.5 mM MgCl2, 340 mM sucrose, 10% glycerol and 1 mM DTT) supplemented with protease and phosphatase inhibitors (Sigma). 2 μL of Triton X-100 (from 5% stock) was then added to the fractionation set, mixed for 3 s using vortex (highest speed) and placed on ice for 5 min. Lysates were then spun at 1300 g for 4 min at 4°C, 80 μL of the supernatant was collected as soluble fraction and mixed with 20 μL of 5× Laemmli buffer. Remaining pellet was gently washed with 1 mL of buffer A (pellet was not disturbed), spun again and re-suspended in buffer C (50 mM Tris-HCl pH7.4, 9 M Urea and 150 mM 2-mercaptoethanol), thoroughly sonicated and spun at 21 000 g. Similarly, 80 μL of the supernatant was collected as insoluble fraction and mixed with 20 μL of 5× Laemmli buffer. The whole lysate set was re-suspended in buffer C, sonicated and spun at 21 000 g for 15 min at 4°C. Again, 80 μL of the supernatant was collected as whole cell lysate sample and mixed with 20 μL of 5× Laemmli buffer. All samples were then boiled for 5 min at 95°C and analyzed by immuno-blotting (same volume of each sample was loaded) with the indicated antibodies.

DNA was extracted from cells (10⁶ cells per sample) using the QIAGEN Core B kit and re-suspended in 20 mM Tris-HCl. 30 ng of genomic DNA was amplified in a PCR reaction containing φ29 polymerase, 1% Tween-20, 200 μg/mL BSA and dTTP, dGTP and dATP for 8 h at 30°C followed by 20 min at 65°C. PCR reactions were run with and without φ29 polymerase to ensure the signal was specific for rolling-circle amplification products. Amplified samples were then blotted onto Zeta-Probe membrane (Bio-Rad) using a slot blotter. DNA was crosslinked to the membrane with a UVA Stratalinker 2,400 and membranes were then soaked in PerfectHyb Plus (Sigma Aldrich) for 20 min at room temperature. A 3’ Digitonin (DIG) tagged [CCCATT]5 oligonucleotide was then diluted in PerfectHyb to a final concentration of 40 nM in 20 mL hybridization buffer, and was hybridized with the membrane for 2 h at 37°C. Following hybridization, membranes were briefly washed twice with wash buffer (0.1 M Maleic acid, 3 M NaCl; 0.1% Tween 20 adjusted to pH7.5), before blocking for 30 min and probing with a DIG antibody for 30 min. Finally, membranes were washed 3 times with wash buffer before placing into a cassette with CDP-Star solution. Blots were then developed onto Amersham Hyperfilm ECL film.

For growth curves, MCF7 and U2OS cells were seeded in duplicate in 96 well plates at 2,500 cells per well (including no cells blank). Each day, separate duplicate wells were incubated with DMEM Fluorobrite Media (A1896701, GIBCO) supplemented with 10% fetal bovine serum, 1% glutamate, 1% penicillin/streptomycin and 10 μg/mL resazurin for 2 h. After the incubation time, fluorescence in the appropriate wells was read using microplate reader (CLARIOStarPlus, BMG LABTECH) with the following settings: excitation λ = 530 nm, emission λ = 590 nm. For the cell survival assays, 24 h post-seeding old media was replaced with fresh media containing the indicated drugs at different concentrations and cells incubated for another 72 h. On the final day the survival plates were assayed identically as growth curve plates. Cell growth was expressed as a relative % of wild-type cells whereas survival of the cells was expressed as relative % of the corresponding untreated samples.

For pull-down of FLAG-tagged proteins, transiently transfected HEK293T (single 10 cm dish per sample) were lysed in pull-down buffer (20 mM HEPES pH7.4, 150 mM KCl, 10% glycerol, 2 mM MgCl2, 0.5 mM DTT, 0.5% NP-40) supplemented with protease and phosphatase inhibitors (Sigma) as well as benzonase (250 U; Sigma) for 1 h at 4°C on a rotating wheel. Lysates were then spun at 21,000 g for 15 min at 4°C and protein concentration determined using Bradford assay (Sigma). Approximately 3–4 mg of total cell extract was mixed with 20 μL of M2-anti-FLAG agarose beads (bed volume) previously washed once with 1 mL of pull-down buffer and incubated for 2 h at 4°C on a rotating wheel. After the incubation, beads were spun (30 s at 2700 g at 4°C), washed 5 times with the pull-down buffer (at the second wash beads were transferred to a fresh tube to prevent contamination with sticky lysate proteins) and eluted with 30 μL of 3×FLAG peptide at 400 μg/mL for 30 min at 4°C shaking. Elutions were then mixed with 5× Laemmli buffer, boiled for 5 min at 95°C and analyzed by immuno-blotting with the indicated antibodies.

Approximately, 10⁶ of U2OS or MCF7 cells were pulsed with 50 μM BrdU for 5 min, harvested by trypsinization and re-suspended in 1 mL of PBS. 2 mL of −20°C absolute ethanol was then added dropwise to the cells while gently vortexing. Cells were fixed for at least 30 min at room temperature, spun for 5 min at 500 g and ethanol removed by two washes with PBS. To reveal BrdU epitopes, samples were treated with 1 mL of 2 M HCl/0.5% Triton X-100 solution for 30 min at room temperature, followed by acid neutralization with 1 mL of 0.1 M sodium tetraborate pH8.5. Cells were then blocked with 1% BSA/0.5% Triton X-100 in PBS for 30 min at room temperature, counted and brought to concentration of 10⁶ cells/mL. 20 μL of the anti-Brdu antibody was added to the cells and incubated for 1 h at room temperature, followed by three washes with 1% BSA/0.5% Triton X-100 in PBS and incubation in 100 μL of anti-mouse Alexa 488 in 1% BSA/0.5% Triton X-100 for 1 h at room temperature. Samples where again washed three times with 1% BSA/0.5% Triton X-100 in PBS and treated with RNase A at 100 μg/mL and propidium iodide at 40 μg/mL in PBS for 30 min at room temperature. Analysis of the samples was performed using SH800z Cell Sorter instrument (SONY).

Chromosome spreads were prepared with standard techniques. Briefly, cells were incubated with Karyomax Colcemid (Thermo Fisher Scientific) 50 ng/mL for 3 h. Following a swelling step in hypotonic (Buffered Hypotonic Solution, Genial Helix), the cells were fixed twice in Carnoy’s fixative (methanol:acetic acid 3:1). The cells suspension was dropped onto clean slides, and air-dried.

The copy number and position of the FAM111B locus in the different cell lines was analyzed by FISH with fosmid G248P81352D4 (WI2-0409G07, hg19 chr11:58,859,240-58,898,785) mapping to FAM111B genomic position, and two BAC constructs flanking the region, RP11-100N3 (hg19, chr11:56481672-56641043) and RP11-286N22, hg19 chr11:61,197,622-61,249,870). The probes were labelled by nick translation, using a commercial kit (Abbott Molecular Nick Translation Kit) incorporating ChromaTide Alexa Fluor 488-5-dUTP (Thermo Fisher Scientific), ChromaTide Alexa Fluor 594-5-dUTP (Thermo Fisher Scientific), or biotin-16-dUTP (Sigma). The probes were re-suspended in hybridization buffer (50% formamide, 10% dextran sulfate, 2× SSC) at 10 ng/µL, in the presence of a 10× excess of unlabeled human Cot1 DNA (Sigma Aldrich). The cellular DNA was denatured in NaOH 0.07 M. The probe DNA was denatured by incubating it at 85°C and allowed to renature at 37°C for 30 min. The hybridization was carried out overnight at 37°C. Following post-hybridization washes in 0.1×SSC at 60°C, the biotinylated probe was visualized using Avidin Cy5 (ThermoFisher Scientific). A minimum of 25 images per cell line were acquired and analyzed with the Leica Cytovision software, on an Olympus) BX-51 epifluorescence microscope equipped with a JAI CVM4+ progressive-scan 24 fps black and white fluorescence CCD camera.

The TeloFISH was carried out using the Telomere PNA FISH Kit/Cy3 (Dako), following the manufacturer instructions. A minimum of 30 metaphases per cell line were acquired, using a Leica DM6B microscope for epifluorescence, equipped with a DFC 9000  Gt B&W fluorescence CCD camera, and operated via the Leica LASX software.

Following the FLAG-IP, elutions were denatured in 200 μL 8 M urea (4.8 g/10 mL) in 100 mM TEAB for 30 min at room temperature, reduced with 10 mM TCEP, for 30 min at room temperature, alkylated with 50 mM iodoacetamide for 30 min at room temperature in the dark and finally diluted down to 1.5 M urea with 50 mM TEAB. Protein digestion was performed using 1.5 ng of trypsin in 50 mM TEAB over night at 37°C, followed by clean up with C18 column (PepMapC18; 300 µm × 5 mm, 5 µm particle size, Thermo Fischer) using solvent A (0.1% formic acid in water) at a pressure of 60bar and separated on an Ultimate 3000 UHPLC system (Thermo Fischer Scientific) coupled to a QExactive mass spectrometer (Thermo Fischer Scientific). The peptides were separated on an Easy Spray PepMap RSLC column (75 µm i.d. × 2 µm × 50 mm, 100 Å, Thermo Fisher) and then electro sprayed directly into an QExactive mass spectrometer (Thermo Fischer Scientific) through an EASY-Spray nano-electrospray ion source (Thermo Fischer Scientific) using a linear gradient (length: 60 min, 5%–35% solvent B (0.1% formic acid in acetonitrile and 5% dimethyl sulfoxide), flow rate: 250 nL/min). The raw data was acquired on the mass spectrometer in a data-dependent mode (DDA). Full scan MS spectra were acquired in the Orbitrap (scan range 380-1800 m/z, resolution 70000, AGC target 3e6, maximum injection time 100 m). After the MS scans, the 15 most intense peaks were selected for HCD fragmentation at 28% of normalized collision energy. HCD spectra were also acquired in the Orbitrap (resolution 17500, AGC target 1e5, maximum injection time 128 m) with first fixed mass at 100 m/z. Progenesis QI (Waters) based label-free quantitation results were imported into Perseus 1.5.2.463. Quantitative data was log2 transformed and normalized by median subtraction and missing values were imputed based on normal distribution.

Image analysis was performed using CellProfiler v. 4.2.1 software and home design pipelines. Briefly, for the measurements of TeloFISH foci intensities, individual chromosome objects were identified using Watershed algorithm and DAPI channel. Neighbouring nuclei were removed using twostep process: step one - Measure Object Size and step two Filter Objects by size (nuclei > chromosomes). Individual chromosomes were then used to mask TeloFISH channel (Mask Image) and telomers identified using Identify Primary Objects algorithm. Finally, intensity of individual TeloFISH signals was determined with Measure Object Intensity algorithm. For the TRF2 foci intensity measurements, nuclei were identified in DAPI channel (Identify Primary Objects) and used as a mask (Mask Image) to identify nuclear TRF2 objects coming from telomeres (Identify Primary Objects). The intensity of TRF2 objects was then measured using Measure Object Intensity algorithm. For FAM111B intensity in PCNA positive and negative cells, DAPI channel was used to identify nuclei which later masked PCNA channel. Intensity of PCNA in each nucleus was determined and a manual threshold was set up to differentiate PCNA positive and negative cells. Filtered PCNA nuclei were then masked onto FAM111B channel and FAM111B intensity quantified in both populations.

For the statistical calculations, Graphpad Prism 9 software was used. Briefly, normal distribution was assumed for experiments where manual counting was involved (Ona-Way Anova, OWA) whereas greater number of datapoints obtained with CellProfiler was analyzed with OWA (Gaussian distribution) or Kruskal–Wallis (K-W, non-Gaussian distribution) tests to determine whether the observed differences between sample groups were significant.