Hat1 Orchestrates Heterochromatin Inheritance by Regulating Localization of H3K9 Methyltransferases

Many regions of heterochroma1n associate with the nuclear periphery and are known as Lamin-associated domains (LADs). Histone acetyltransferase 1 (Hat1) is a highly conserved enzyme which acetylates newly synthesized histones H4 on lysines 5 and 12 prior to their deposi1on on chroma1n. Hat1 is required to preserve chroma1n accessibility within a subset of LADs called Hat1-dependent accessibility domains (HADs). Here we profile a diverse set of histone modifica1ons in Hat1 KO and WT immortalized mouse embryonic fibroblasts (iMEFs) and find that Hat1 regulates diverse aspects of the structure of HADs and non-HAD LADs (nhLADS). In HADs, these changes include the conversion of H3K9me2 to H3K9me3. Analysis of H3K9-specific histone methyltransferases (HMTs) shows that that Suv39h1 and Suv39h2 have dis1nct localiza1on paTerns, where only Suv39h2 localizes to LADs. G9a only localizes to LADs in regions enriched for H3K9me2. We find that Hat1 loss results in a redistribu1on of these HMTs in both HADs and nh LADs. There is a decrease in the levels of G9a with a concomitant increase in Suv39h2. These results suggest Hat1 func1ons to restrain the forma1on of a more strongly heterochroma1c state and highlight a role for Hat1 as an essen1al regulator of heterochroma1n inheritance. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 4 Introduc.on Eukaryo1c chroma1n can be broadly divided into euchroma1n and heterochroma1n. These two classes of chroma1n perform different func1ons, localize differently in the nucleus, and are established and inherited by different mechanisms (1-4). Euchroma1n is gene rich, transcrip1onally ac1ve, highly acetylated, and tends to localize to the interior of the nucleus, while heterochroma1n is hypoacetylated, gene poor, and enforces transcrip1onal silencing through physical compac1on and recruitment of repressive transcrip1on factors (5,6). Heterochroma1n falls into two main classes. Faculta1ve heterochroma1n is characterized by H3K27me3 and H2AK119ub and primarily silences genes in a developmental or cell cycle-specific manner(7-9). Cons)tu)ve heterochroma)n is characterized by H3K9me2/3 and promotes long-term silencing of gene-poor regions, transposons, and structural elements like centromeres and telomeres(10). These histone PTMs can also recruit heterochroma1n-promo1ng factors such as the H3K9 methyla1on-binding protein HP1, which promotes chroma1n compac1on(11,12). Both classes of heterochroma1n are depleted in histone acetyla1on and chroma1n accessibility, but these features are par1cularly pronounced in cons1tu1ve heterochroma1n. One way the cell maintains the structural and func1onal separa1on between euchroma1n and heterochroma1n is by tethering the laTer to the nuclear periphery. Lamin-Associated Domains (LADs) are large chroma1n domains (ranging from 10s of kilobases to megabases) which associate with the nuclear lamina by interac1ons with lamins and lamin-associated proteins (13). LADs contain a large frac1on of the cons1tu1ve heterochroma1n in the genome and are enriched in H3K9me2/3 and HP1. For example, HP1 promotes lamina associa1on by binding histones and interac1ng with lamina-associated proteins like LBR and PRR14 (14,15). Other histone modifica1ons, such as H3K9me2, H3K9me3, and H3K27me3 also influence lamina- interac1ons (13,16). Proper regula1on of chroma1n-lamina associa1on is essen1al at both the cellular and organismal levels. Many LADs experience dynamic changes in lamina associa1on through differen1a1on and development and vary across cell types, earning the label of faculta1ve LADs (fLADs), in contrast to cons1tu1ve LADs (cLADs), which are always lamina- associated across cell types (17,18). Mis-regula1on of lamina associa1on can influence many features and processes, including DNA replica1on, genome organiza1on, transcrip1onal regula1on, differen1a1on, and cell fate commitment (19-26). Regula1on of LAD chroma1n is complex. Despite its repressive character, lamina associa1on does not always correlate neatly with transcrip1onal silencing and LADs can contain low levels of ac1ve features like histone acetyla1on, oben found in discrete peaks near sparse sites of accessible chroma1n. During S-phase, a cell must successfully replicate both its genome and epigenome. As the replica1on fork moves along the DNA, nucleosomes are disassembled from the parental strand .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 5 and H3-H4 tetramers are recycled by replisome-associated chaperones onto the daughter strands near their original loca1ons (1,27). Chroma1n assembly factor 1 (CAF-1) is recruited to the replica1on fork by interac1ons with PCNA and delivers newly synthesized histones to restore nucleosome levels (28,29), resul1ng in roughly equal propor1ons of new and parental histones on the daughter strands. Many of these new histones are deposited already bearing PTMs, most notably acetyla1on of histone H4 lysines 5 and 12. Following DNA replica1on, nucleosome organiza1on is restored by chroma1n remodelers and histone methyltransferases working to copy parental heterochroma1n marks onto the new histones. Six different histone methyltransferases (HMTs) target H3K9 in mammals (30,31). Of these, Suv39h1/Suv39h2 primarily deposit H3K9me3 and G9a/GLP primarily deposit H3K9me2. Setdb1 can mono-, di-, or tri-methylate H3K9 (30,32). H3K9me1 deposited by Setdb1 helps prime nascent chroma1n for further methyla1on by Suv39h1/2, and Setdb1 oben contributes H3K9me3 at repe11ve elements like endogenous retroviruses (ERVs) (33-36). Suv39h1/2 are homologs of the fission yeast HMT Clr4, which binds H3K9me3 through its chromodomain and deposits the same mark onto nearby histones in a “read-write mechanism”(37-39). This same mechanism has been demonstrated in Suv39h1 and is presumed to be shared by Suv39h2, though Suv39h1 is generally treated as the primary H3K9 trimethylase of the two (40). Faculta1ve heterochroma1n is restored in a similar manner by coopera1on between the H3K27me3 writer PRC2 and PRC1-dependent H2AK119ub (41-43). Numerous factors may influence the efficiency of the read-write mechanism, including the density of the mark in ques1on , the 3D organiza1on of chroma1n, the presence of other modifica1ons, linker histones, and chroma1n binding proteins like HP1 (39,44-49). However, the restora1on of H3K9me3 and H3K27me3 takes place following S-phase very slowly and much work remains to understand how their inheritance is regulated (27,50,51). Histone acetyltransferase 1 (Hat1) binds newly synthesized H3-H4 dimers to acetylate histone H4 on lysines 5 and 12 before the histones are deposited onto nascent DNA by CAF-1 (52-54). Hat1 is highly conserved across eukaryotes, and its importance seems to scale with organismal complexity (53,55-57). While Hat1 dele1on causes only mild phenotypes in yeast, dele1on in Drosophila leads to wide-spread gene mis-regula1on during development, and Hat1 KO in mice is embryonic or neonatal lethal(57,58). However, the main molecular func1on of Hat1 remains unclear. In addi1on to acetyla1ng nascent H4, Hat1 chaperones histones and can directly associate with nascent chroma1n, though it is unable to acetylate nucleosomal histones(54,59- 61). Both its associa1on with chroma1n and H4K5/12ac are short-lived, disappearing from chroma1n within 2 hours aber replica1on (61-63). However, the placement of H4K5/12ac on nascent chroma1n leaves them perfectly posi1oned to influence the early stages of chroma1n matura1on. Consistent with this idea, we recently demonstrated that loss of Hat1 has a significant impact on the chroma1n landscape. Heterochroma1n domains ranging from tens of .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 6 kilobases to megabases in size displayed a loss of chroma1n accessibility upon Hat1 KO, and were therefore called Hat1-dependent accessibility domains (HADs)(64). HADs are a subset of LADs and thus share their chroma1n characteris1cs, including enrichment of H3K9 methyla1on, poor chroma1n accessibility, and low gene density. Furthermore, nascent chroma1n in Hat1 KO cells is enriched in cons1tu1ve heterochroma1n associated proteins and depleted in histone acetyla1on rela1ve to nascent chroma1n from Hat1 WT cells, sugges1ng that these chroma1n changes arise from loss of Hat1’s S-phase ac1vity (61). Here we set out to more fully understand the effects of Hat1 loss on the chroma1n landscape. We find that Hat1 regulates the levels of numerous ac1ve and repressive features across LADs, and especially within HADs. HADs require Hat1 to preserve the epigene1c signature of unannotated sites which resemble primed enhancers and to maintain H3K9me2 and prevent excessive accumula1on of H3K9me3 and HP1β. While other LADs do not accumulate these cons1tu1ve heterochroma1n features upon Hat1 loss, they s1ll require Hat1 to preserve both

histone acetyla1on and H3K27me3, a mark of faculta1ve heterochroma1n which marks LAD borders and regulates chroma1n-lamina interac1ons. We trace the changes in H3K9 methyla1on to redistribu1on of H3K9 HMTs and show that Suv39h2, not Suv39h1, is primarily responsible for H3K9me3 in LADs. In addi1on, H4K5ac increases at LAD boundaries upon Hat1 loss, sugges1ng a mechanism by which the cell protects the surrounding genome from invasion of heterochroma1n. Finally, proximity labeling demonstates that Hat1 is in proximity to nuclear lamina components, which is confirmed by proximity labeling assays with Hat1 and lamin B. These results show Hat1 to be an essen1al regulator of heterochroma1n inheritance, most prominently in domains associated with the nuclear lamina.

Cell culture Mouse embryonic fibroblasts harvested from Hat1 WT and KO mice and immortalized as previously described were cultured at 37 ⁰C in 5% CO2 (58). Cells were grown in Dulbecco’s Modified Eagle’s High Glucose Medium supplemented with 10% FBS, 1% glutamine, and 1% Penn-Strep and passaged approximately every 2-3 days with 0.02% Trypsin. Western Blot Whole cell lysate was prepared in NP-40 Lysis Buffer and protein concentra1on was quan1fied by Nanodrop. Following electrophoresis, protein was transferred from the gel to a nitrocellulose membrane using a Bio-Rad Trans-Blot Turbo at 25 V for 20 minutes. The quality of the run and transfer was checked by visualizing the protein with Ponceau stain, which was then rinsed off with TBS-T. Membranes were blocked for 1 hr in 5% milk in TBS-T, rinsed with TBS-T, and incubated with primary an1body in 2.5% milk at 4 ⁰C overnight. Membranes were rinsed 3X for .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 7 5 min each in TBS-T, then incubated for 1 hr with HRP-conjugated secondary an1body diluted 1:5000 in 1% milk. Membranes were rinsed again 3X for 5 min in TBS-T. Pierce ECL Western Bloong Substrate (Thermo Scien1fic) was applied to the membrane, which was visualized using a Sapphire Biomolecular Imager (Azure Biosystems). CUT&Tag CUT&Tag was performed as previously described (65) according to the EpiCypher® CUTANA™ Direct-to-PCR CUT&Tag Protocol, with a light crosslinking step adapted from (66). Briefly, 100,00 iMEFs were harvested and rinsed with PBS, then incubated in Nuclear Extrac1on Buffer for 10 minutes on ice. Nuclei were spun down, resuspended in PBS, and fixed by adding formaldehyde to a concentra1on of 0.1% and incuba1ng 2 min at room temperature. The formaldehyde was quenched with two molar equivalents of glycine. Cells were spun and resuspended in Nuclear Extrac1on Buffer, then combined with 10 μL ac1vated Concavalin A magne1c beads and incubated for 10 minutes at room temperature for nuclei to bind to the beads. The beads were separated with a magnet, resuspended in An1body 150 Buffer with the appropriate dilu1on of primary an1body, and incubated overnight at 4 ⁰C. The primary an1body solu1on was removed and nuclei were incubated for 30 min with secondary an1body. Nuclei were rinsed twice with Digitonin 150 Buffer, then incubated 1 hr with pAG-Tn5 fusion transposase. Next, they were rinsed twice with Digitonin 300 Buffer and then incubated at 37 ⁰C for 1 hour with 10 mM MgCl2 to ac1vate the Tn5 Transposase. Tagmenta1on was stopped by rinsing with TAPS Buffer, then samples were suspended in SDS Release Buffer and incubated at 58 ⁰C for 1 hr. SDS was quenched in 0.5% TritonX-100. DNA libraries were amplified using CUTANA High-Fidelity 2X PCR Master Mix™ with Universal i5 primer and the appropriate barcoded i7 primer, and DNA was extracted using 1.3X AMPure beads. Libraries were sequenced with paired-end Illumina sequencing. CUT&RUN CUT&RUN was performed according to the EpiCypher® CUT&RUN Protocol. 500,000 iMEFs were harvested and rinsed with PBS, suspended in Nuclear Extrac1on Buffer (as with CUT&Tag) for 10 min on ice, rinsed with Wash Buffer, bound to ac1vated Concavalin A beads, and incubated with primary an1body overnight at 4 ⁰C (Suv39h1 - Abcam ab283262; Suv39h2 - Abcam ab190870; G9a - Cell Signaling 68851; Setdb1 - Proteintech 11231-1-AP, Lamin B1 - Abcam ab16048). For Lamin B1 and IgG, nuclei were not extracted. 500,000 iMEFs were harvested, rinsed twice with wash buffer, bound to ac1vated Concavalin A beads, and then incubated overnight with primary an1body at the appropriate dilu1on in An1body Buffer. Samples were rinsed twice with Digitonin Buffer and incubated for 10 min with pAG-MNase. There were then rinsed twice more with Digitonin Buffer and incubated in 2 mM CaCl2 for 2 hrs at 4 ⁰C. Samples were quenched with 33 μL STOP Buffer containing E. coli spike-in DNA and incubated 10 min at 37 ⁰C to release .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 8 fragmented DNA, then beads were separated out of solu1on on a magne1c rack and the supernatant was transferred to new tubes. Libraries were prepared for next genera1on sequencing using the CUTANA CUT&RUN Library Prep Kit. ChIP-seq and RNA-seq ChIP-seq was performed as previously described (64). RNA-seq was performed in biological triplicate as previously described (67). Processing Sequencing Data Raw sequencing data was analyzed to the mm10 genome with parameters --end-to-end --very- sensi-ve --no-unal --no-mixed --no-discordant --phred33 -I 10 -X 700 and filtered to remove mitochondrial reads and those mapping to unlocalized or unplaced con1gs. All Hat1 WT and KO libraries belonging to the same an1body/target were normalized to one another by library size. The E. coli spike-in which had been added to some CUT&RUN libraries was ignored as it did not yield consistent scaling between replicates. Scaling factors for each sample were calculated by dividing the number of filtered, mapped reads in the sample with the fewest mapped reads by the number of mapped reads in each sample. Scaling factors reads for each sample were calculated by dividing the number of filtered, mapped reads in the sample with the fewest mapped reads by the number of mapped reads in each sample. Library sizes were adjusted by subsampling with samtools view -s. Biological replicates were merged for downstream analysis and viewing using samtools merge. The normalized libraries were then converted to bedGraph and or bigwig format for viewing on IGV and other downstream analyses. For histone PTMs, tracks showing the Hat1 KO/WT log2-fold change across the genome were created by combining the total signal for that PTM into bins across the genome using bedtools. The size of the bins varied depending on the mark and ranged from 10 kb for more diffuse or abundant marks to 300 kb for sparse marks. The WT and KO data was then loaded in R. The log2 KO/WT values were calculated for each bin, the data was filtered to remove bins with undefined log2FC values and very low signal, and a background level of reads was imputed to remaining empty bins with undefined logFC. The log2FC values were then recalculated across the genome. For some marks, a 70 kb sliding window average was used to smooth out the track and make domains clearer. For logFC tracks for CUT&RUN datasets of HMTs were created more simply using deepTools bigwigCompare with 30 kb bins. chromHMM To run chromHMM on the WT and KO data together, a chromosome sizes file with “WT_” appended to the beginning of each chromosome name was combined with an iden1cal file that had “KO_” added to each chromosome name. We went through each bam file and added “WT_” or “KO_” to the beginning of all chromosome names in each WT and KO file, respec1vely. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 9 Paired WT and KO bams were then combined using samtools merge so that they could be treated as an individual sample but would map to separate chromosomes due to the modified chromosome names and could be re-separated later. Binarized files were created with 10 kb bins and the chromHMM algorithm was run for a range of states numbers. We chose to proceed with 10 states. Aber running the algorithm, the dense bed file was divided into two files based on WT_ and KO_ chromosome prefixes and the prefixes were removed from each file. These were used for downstream analysis in R and visualiza1on with IGV. For analysis in R, a track with all gaps of unknown nucleo1des downloaded from UCSC table browser was used to remove all bins had 50% or greater overlap with unmappable regions. All plots were created with ggplot2. Sankey plots were created using the ggsankey package. Data Analyses All LAD analysis was performed using our previously published set of LADs mapped in the same cells used for this study (65). Sex chromosomes were excluded from all analyses. For barplots of H3K4me1 (Supplemental Code S2.1) and H3K27ac (as in Supplemental Code S2.1) BED files were loaded into R and the number of peaks which overlapped the regions of interest (i.e. HADs or nhLADs) were summed for each of the 6 biological replicates (3 WT and 3 Hat1 KO). The averages across replicates for each group were ploTed in ggplot2. Student’s t-test was used to calculate the p-values. To make LAD heatmaps and profile plots, a matrix of all marks scaled across LADs was generated using deepTools computeMatrix and loaded in R (Supplemental Code S2.2). Heatmaps were made with the pheatmaps package and profile plots were made with ggplot2, except the profiles/heatmaps in Figure S1, which were made with deepTools plotHeatmap. To calculate the frac1on of lamina-associated chroma1n changed for each feature, we combined the signal into 10 kb bins across the genome, calculated the log2 KO/WT fold-change for each bin, and selected all bins which had at least 5 kb (50%) overlap with a LAD. Bins with log2FC > log2(1.25) were considered to have gained the feature and those with log2FC < -log2(1.25) were considered to have lost it. To determine the net change of a given feature in LADs, we calculated the total signal for each LAD using bedtools map and calculated the log2 KO/WT fold-change for each LAD. LADs with log2FC > log2(1.25) were considered to have gained the feature and LADs with log2FC < - log2(1.25) were considered to have lost it. Regression analysis was performed in R and ploTed using ggplot2. The r value is the Peason correla1on coefficient and the p-value is the significance of the slope. Regression was performed in 3 different ways. (1) Abundance of feature 1 in 10 kb bins versus abundance of feature 2 in 10 kb bins. Regression was performed using all bins within LADs, all bins within .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 10 HADs, and all bins in non-HAD LADs. Outliers were removed using a more stringent version of Tukey’s rule: bins were dropped in which the log-transformed value fell more than 4*IQR above the 3rd quar1le or more than 4*IQR below the 1st quar1le. Due to the larger number of points, the plots depict the density of points across a 2-dimensional grid. (2) Total abundance of feature 1 per LAD versus total abundance of feature 2 per LAD. Outliers were not dropped. Each point in the plot represents a LAD. (3) log2 KO/WT fold-change of feature 1 in LADs versus log2 KO/WT fold-change of feature 2 in LADs. Outliers were not dropped. Each point in the plot represents a LAD. APEX2 Transfec/ons HEK293T cells were cultured in in DMEM media (Sigma) supplemented with 10% FBS (Sigma) and 1X Penicillin/Streptomycin an1bio1cs (Gibco). Constructs containing pcDNA-dest40 vector and APEX2-HAT1 fusion or APEX2 only were made using the Gateway Mul1site Cloning kit (Thermo Fisher). Cells were transfected with the constructs at ~80% confluency using Lipofectamine 2000 (Life Technologies) as per manufacturer’s protocol. APEX2 Labeling 24 hours post-transfec1on cells were labeled as described with minor modifica1ons [82]. Briefly, cells were incubated in 500uM bio1n-phenol for 30 min at 37°C. Next, the cells were labeled for 1 min by addi1on of H2O2 to the final concentra1on of 1mM. Labeling reac1on was quenched by freshly prepared quencher solu1on (10 mM sodium ascorbate, 5 mM Trolox, and 10 mM sodium azide in PBS). Cells were collected and lysed in RIPA buffer (50mM Tris-HCl, pH 7.5, 150mM NaCl, 1% NP-40, 1mM EDTA, 1mM EGTA, 0.1% SDS, and 0.5% sodium deoxycholate supplemented with 1X Complete Protease inhibitor cocktail (Roche), 1mM PMSF, 10 mM sodium azide, 10 mM sodium ascorbate, and 5 mM Trolox). To help clarify lysates, cells were sonicated with Diagenode Bioruptor Sonicator on high seong for 10 min with 30sec on/off. Addi1onally, cells were treated with Benzonase nuclease (Sigma) and incubated for 1 hour at 4°C with rota1on and consequently centrifuged for 20 min at 20000g. IP and Mass Spectrometry APEX2-MS experiment was performed in triplicate. HEK293T cells were transfected with APEX2- HAT1 fusion construct, APEX2-only construct, and untransfected cells were used as a control. APEX2-HAT1-transfected cells, APEX2-only transfected cells, and untransfected cells were treated with either bio1n-phenol+H2O2 or bio1n-phenol alone. Cells were labeled, and cell lysates were prepared as described above. Bio1nylated proteins were isolated with streptavidin magne1c beads (Thermo Scien1fic). Samples were incubated with beads overnight with rota1on at 4°C. Aber rota1on, beads were captured and washed once with RIPA buffer, once with 1M KCl, once with 0.1M sodium carbonate, once with 2M urea in 10mM Tris-HCl, pH 8.0, and twice with RIPA lysis buffer. Samples were digested with trypsin via on-bead diges1on overnight at 37°C and and supernatant dried by vacuum concentrator. Liquid chromatography- .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 11 nanospray tandem mass spectrometry (Nano-LC/MS/MS) was performed on an Orbitrap Fusion mass spectrometer (Thermo Scien1fic) operated in posi1ve ion mode. Pep1des (1 µg) were separated on an easy spray nano column (PepmapTM RSLC, C18 3µ 100A, 75µm X150mm, Thermo Scien1fic) using a 2D RSLC HPLC system (Thermo Scien1fic) at 55°C . Each sample was injected into the µ-Precolumn Cartridge (Thermo Scien1fic) and desalted with mobile phase A (0.1% formic acid in water) for 5 minutes. Flow rate was set at 300nL/min and pep1des eluted with increasing mobile phase B (0.1% formic acid in acetonitrile) over 109 min as follows: 2% to 20% in 60 min, 20-32% in 15 min, from 32-50% in 10 min, 50-95% in 5 min (holding at 95% for 2 min) and back to 2% in 2 min. The column was equilibrated in 2% of mobile phase B for 15 min before the next sample injec1on. APEX2 Data Analysis Mass spectra from all RAW data files were converted to mzML with ProteoWizard and OpenMS (v 2.5.0(68,69). Converted files were searched on the OpenMS plaorm with MSGF+ search engine against a reviewed UniProt human proteome (downloaded 09/24/2020) containing the cRAP and MaxQuant contaminant FASTAs. Search parameters included: full trypsin digest, 1 missed cleavage, oxida1on of methionine and acetyla1on of lysine as variable modifica1ons, precursor mass tolerance 20 ppm and fragment mass tolerance 0.8 Da. Tes1ng for differen1ally expressed proteins was performed in R with the limma package version 3.42.2 using an imputa1on method described in Gardner and Freitas (70). Briefly, samples were selected for pair-wise comparisons prior to filtering out lowly expressed proteins, and missing values were imputed with a mul1ple imputa1on approach by treatment group. Data was quan1le normalized, and significance (p-value < 0.05) determined by a modified exact test. To correct for background bio1nyla1on, significantly bio1nylated proteins were iden1fied in APEX2-Hat1 bio1n-phenol+H2O2 (BPH2O2) condi1on compared to APEX2-Hat1 bio1n-phenol only (BP) condi1on. Next, significantly bio1nylated proteins were iden1fied in APEX2-only BPH2O2 condi1on compared to APEX2-only BP condi1on. Also, significantly bio1nylated proteins were iden1fied in untransfected BPH2O2 condi1on compared to untransfected BP condi1on. Finally, proteins found to be significantly bio1nylated in the two laTer condi1ons were removed from the list of proteins significantly bio1nylated with APEX2-Hat1 construct. Proximity Liga.on Assays Three Hat1 WT or Hat1 KO MEF cell lines were seeded in equal quan11es on coverslips and allowed to aTach for 24 h. Cells were then permeabilized with 0.5% Triton X-100 and fixed with 4% PFA simultaneously for 15 min, rinsed with PBS, and fixed again with 4% PFA for 10 min at room temperature. Aber several PBS washes cells were blocked with 5% BSA for 1 h at room temperature. BSA was removed with PBS and primary an1bodies detec1ng two proteins of interest were diluted in 1% BSA, 0.3% Triton X-100 and added to cells overnight at 4 °C (mouse an1-LaminB1, 1:200, Abcam ab8982; rabbit an1-LaminB1, 1:200, Abcam ab65986; rabbit an1- HAT1, 1:1000, Abcam ab193097). The following day, primary an1bodies were removed with .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 12 PBS and cells were subjected to the DuolinkTM Proximity Liga1on Assay protocol according to the manufacturer’s instruc1ons (Sigma: DUO92008, DUO92004, DUO92002, and DUO82049). Aber amplifica1on, nuclei were stained with 20 mM Hoechst 33342 Fluorescent Stain and mounted on slides using Vectashield. Slides were analyzed under a Zeiss LSM 900 Airyscan 2 Point Scanning Confocal microscope. Images were acquired using Zen Blue 3.0 and quan1fica1on was completed using ImageJ version 1.52t according to a previously described protocol (71). Table 2.1. An.bodies used in experiments An.gen Company Catalog Number Applica.on HP1β Cell Signaling 8676 ChIP-seq, Western Blot H3K9me3 Abcam ab8898 CUT&Tag H3K9me2 Abcam ab1220 CUT&Tag H3K27me3 Cell Signaling 9733 CUT&Tag H3K4me1 Cell Signaling 5326 CUT&Tag H3K9ac Cell Signaling 9649 CUT&Tag H3K14ac Abcam ab52946 CUT&Tag H3K27ac Cell Signaling 8173 CUT&Tag H4K5ac Abcam ab51997 CUT&Tag H4K12ac Abcam ab46983 CUT&Tag Lamin B1 Abcam ab16048 CUT&RUN Lamin B1 Abcam ab8982 PLA Lamin B1 Abcam ab65986 PLA Suv39h1 Abcam ab283262 CUT&RUN, Western Blot Suv39h2 Abcam ab190870 CUT&RUN, Western Blot G9a Cell Signaling 68851 CUT&RUN, Western Blot Setdb1 Proteintech 11231-1-AP CUT&RUN, Western Blot IgG Epicypher 13-0042 CUT&Tag, CUT&RUN HP1α Cell Signaling 2616 Western Blot HP1γ Invitrogen MA3-054 Western Blot Β-Ac1n Cell Signaling 3700 Western Blot GAPDH Cell Signaling 5174 Western Blot Hat1 Abcam ab193097 Western Blot, PLA

Hat1 Regulates Accessibility at Primed Enhancers in Heterochroma.n .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 13 Previous results showed that Hat1 was required to preserve peaks of chroma1n accessibility within a subset of Lamin-associated chroma1n that were termed Hat1-dependent accessibility domains (HADs). Given the highly repressed and condensed nature of cons1tu1ve heterochroma1n, these peaks of accessibility are sparsely distributed and their role is unclear. Comparison with publicly available datasets suggested that they overlap with histone modifica1ons characteris1c of ac1ve enhancers (H3K4me1 and H3K27ac). However, they do not coincide with previously known enhancers (64). To test whether the heterochroma1c peaks of accessibility overlap with these enhancer marks in our cell lines and whether those modifica1ons are regulated by Hat1, we performed CUT&Tag in immortalized mouse embryonic fibroblasts (iMEFs) to detect the general enhancer mark H3K4me1 and the ac1ve enhancer mark H3K27ac (72). Just as we saw previously with chroma1n accessibility, we observed a net loss of H3K4me1 peaks in Hat1 KO HADs compared to WT HADs (Figure 1A). The effect of Hat1 on H3K4me1 was specific to HADs as there was no overall loss of H3K4me1 in non-HAD LADs (nhLADs) or in inter-LADs (Figure 1B and Figure S1A,B). While some peaks of H3K27ac were also lost, we did not see a net reduc1on of H3K27ac peaks across HADs, nhLADs, or inter-LADs (Figure 1B and Figure S1A,C). Biological variability between replicates from different mice prevented accurate quan1fica1on of the overlap between H3K4me1 peaks and lost ATAC-seq peaks, but visual comparison showed that many of the lost peaks of accessibility did overlap with peaks of H3K4me1, while a much smaller number also overlapped with H3K27ac. Figure 1C shows a genome browser view of a 175 kb region of a HAD on chromosome 1 that contains the Serpin B8 gene. Consistent with the designa1on of this region as a HAD, there were dis1nct peaks of ATAC-seq signal that were present in Hat1 WT cells that were absent in Hat1 KO cells. The most prominent of these peaks was located just upstream of the Serpin B8 gene. Notably, this Hat1-dependent site of accessibility co-localized with a peak of H3K4me1 that was also Hat1-dependent. While H3K27ac was present upstream of Serpin B8, the signal was more dispersed and was not as obviously Hat1-dependent. Consistent with the hypothesis that Hat1 can regulate heterochroma1c enhancers, RNA-seq analysis indicated that there was a significant Hat1-dependent decrease in Serpin B8 mRNA. Hat1 preserves basal acetyla.on across LADs Hat1 clearly impacts sites of chroma1n accessibility that possess enhancer characteris1cs in cons1tu1ve heterochroma1n. To determine whether the effect of Hat1 is restricted to these discrete loci or whether Hat1 more broadly regulates the structure of chroma1n domains associated with the nuclear lamina, we determined the genome-wide distribu1on of a wide variety of histone PTMs in Hat1 WT and Hat1 KO iMEFs. We performed CUT&Tag for several sites of acetyla1on on histones H3 and H4. These included H4K5ac and H4K12ac, which are direct targets of Hat1, as well as H3K9ac, H3K14ac, H3K27ac. Note that H4K5/12ac can also be .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 14 deposited in a non-replica1on-dependent manner by histone acetyltransferases other than Hat1. Calcula1on of the Hat1 KO/WT log2 fold-change showed a consistent loss of histone acetyla1on across not only HADs, but almost all lamina-associated domains (Figure 2A). The excep1on to this rule was H3K14ac. This modifica1on showed a significant reduc1on in HADs and a more variable response across nhLADs. To visualize the levels of each site of histone acetyla1on across all HADs and nhLADs, we scaled each HAD/nhLAD to a uniform size and ploTed the average signal for each acetyla1on site across them (plus 300 kb up- and down-stream of the HAD/nhLAD borders). For the HAD plots, we used the set of LADs which overlap HADs rather than our previously determined HAD set itself (64). This was for beTer visualiza1on and clearer comparison. The boundaries of LADs can be mapped more precisely than HADs since the increased signal density in domains of lamina associa1on are more easily measured than loss of sparse accessibility peaks. For the rest of this paper, these HAD-overlapping LADs will simply be referred to as HADs. The HAD/nhLAD profiles are shown in Figure 2B. In Hat1 WT cells, H4K5ac, H4K12ac, H3K9ac, and H3K27ac were present at uniformly low levels across the en1rety of both HADs and nhLADs. The levels of these sites of acetyla1on rose drama1cally at the border of the HADs/nhLADs and peaked just outside the HAD/nhLAD borders. Despite this rela1vely low abundance in HADs/nhLADS, loss of Hat1 resulted in a significant decrease in H4K5ac, H4K12ac, H3K9ac, and H3K27ac across the length of HADs/nhLADs (Figure 2B). The H4K12ac, H3K9ac, and H3K27ac profiles changed only within LADs. The peak observed just outside the HAD/nhLAD border was indis1nguishable between Hat1 WT and Hat1 KO cells. In contrast, In Hat1 KO cells there was a marked increase in the level of H4K5ac just outside of HAD/nhLAD borders Figure 2B and 2C). As we recently reported, the paTern observed for H3K14ac was quite dis1nct and differen1ates HADs from nhLADs. In Hat1 WT cells, there was a large peak of H3K14ac just inside the HAD borders that was much higher than the average levels of H3K14ac seen throughout euchroma1n. The level of H3K14ac then gradually decreases throughout HADs to levels somewhat lower than those seen in euchroma1n. In nhLADs, the peak just inside the nhLAD border is maintained but there is a much smaller decrease in the average level of H3K14ac across the length of the nhLAD, with the result that the overall level of H3K14ac across nhLADs was higher than the average levels seen in euchroma1n. In HADs, loss of Hat1 resulted in a decreased level of H3K14ac across the en1rety of the HAD. However, in nhLADs there was no change in H3K14ac. This supports the classifica1on of HADs as a dis1nct set of lamina- associated regions of the genome that are specifically regulated by Hat1. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 15 Very liTle change in histone acetyla1on was seen in euchroma1c areas, indica1ng that Hat1 primarily regulates chroma1n associated with the nuclear lamina. 948 LADs (84.1%), or over 95% of lamina-associated chroma1n, experienced reduc1on of one or more acetyla1on marks upon Hat1 loss. H3K9ac and H3K27ac were the most depleted, being reduced in 64.0% and 72.5% of LADs, respec1vely, while H4K5ac and H4K12ac were lost in 54.0% and 45.8% of LADs, respec1vely. Looking specifically at HADs, the change was even greater, with 173 out of 180 HADs experiencing reduc1on of one or more acetyla1on mark. The reduced histone acetyla1on in Hat1 KO LADs is unlikely to be caused by reduced accessibility of Hat1 KO chroma1n to the pAG-Tn5 transposase, since a control with IgG in place of primary an1body shows only very minor loss of background signal in the absence of Hat1, even in LADs. HADs did show a slight reduc1on in IgG signal in Hat1 KO cells, but it is not nearly sufficient to explain the reduc1on in histone PTM signal observed (Figure S2A,B). Hat1 selec.vely regulates cons.tu.ve heterochroma.n components in HADs and nhLADs The analysis of the impact of Hat1 loss on global histone acetyla1on suggests that Hat1 regulates the chroma1n state of most or all LADs, not only the subset iden1fied as HADs. Informed by previous ChIP-seq data which had shown that Hat1 KO increased the density of H3K9me3 in HADs (64), we hypothesized that the loss of histone acetyla1on in LADs was a downstream result of increased cons1tu1ve heterochroma1n features leading to a more repressive chroma1n state in LADs. To gain a more comprehensive picture of the role of Hat1 in regula1ng repressive chroma1n structure in HADs and nhLAD, we used CUT&Tag to map H3K9me2, H3K9me3, and H3K27me3 in Hat1 WT and Hat1 KO cells. While H3K9me2 and H3K9me3 are thought to characterize cons1tu1ve heterochroma1n and H3K27me3 is linked to faculta1ve heterochroma1n, all of these repressive histone methyla1ons are associated with lamina-associated domains. In fact, LADs can be divided into 3 dis1nct clusters based on which is the dominant methyla1on in a given LAD (Mar1n et al, 2025). Comparing the paTerns of methyla1on between HADs and nhLADs iden1fied several Hat1- dependent altera1ons that differ between HADs and LADs (Figure 3A). As expected, profiles of H3K9me3 show that it was highly enriched in both HADs and nhLADs. H3K9me3 formed a broad plateau spanning most of the HAD or nhLAD, then dropped sharply inside the HAD/nhLAD borders. The level of H3K9me3 in HADs increased in Hat1 KO cells while it slightly decreased in nhLADs, consistent with our previous results (64). Though also characteris1cally enriched in LADs, H3K9me2 displayed a very different profile. H3K9me2 enrichment peaks just inside both HAD and nhLAD borders. However, while loss of Hat1 has liTle net effect on H3K9me2 in nhLADs, there is a drama1c Hat1-dependent loss of H3K9me2 in HADs. As reported previously, H3K27me3 was found in a sharp peak at the border of both HADs and nhLADs. Levels of .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 16 H3K27me3 were lower throughout the LAD interior in both HADs and nhLADs, especially the former. In both groups, loss of Hat1 resulted in reduced H3K27me3 levels across the domains. We saw a notable reduc1on of H3K27me3 in about 75% of LADs, sugges1ng that Hat1 plays a fundamental role in regula1ng H3K27me3 throughout lamina-associated chroma1n. Notably, the changes in histone methyla1on were not strictly confined to HADs (Figure S3A). For example, many nhLADs also gained H3K9me3, but more oben they lost it, resul1ng in a net reduc1on across nhLADs. Viewing H3K9me3, H3K9me2, and H3K27me3 on a genome browser highlights the dis1nct chroma1n structures in HADs and nhLADs and suggests that the opposite effects of Hat1 loss on H3K9me2 and H3K9me3 levels may be linked. Figure 3B shows a ~5 Mb region of chromosome 3 that contains 3 LADs, one of which overlaps a HAD. In the nhLADs, no Hat1-dependent changes in H3K9me3 or H3K9me2 were observed. H3K27me3 was present at higher levels in the nhLADs than the HAD, and experienced a Hat1-dependent decrease in H3K27me3 across both. The HAD displayed a clear Hat1-dependent increase in H3K9me3 that occurs in the same region where there is a Hat1-dependent loss of H3K9me2. The opposing effects of Hat1 on H3K9me2 and H3K9me3 can be observed across all LADs, as a regression analysis of the log2 fold change of these marks in Hat1 WT and Hat1 KO cells shows a marked nega1ve correla1on (Figure 3C). As the primary structural component of cons1tu1ve heterochroma1n that binds to H3K9me2/3, we wanted to determine whether the associa1on of HP1 with HADs and nhLADs is influenced by Hat1. We used ChIP-seq to measure the genome-wide distribu1on of HP1b. We generated profiles of the average signal of HP1b across HADs and nhLADs as done for the histone PTMs. As shown in Figure 3D, there is an increase in HP1b in HADs but a small decrease in HP1b in nhLADs, consistent with the HAD-specific increase in H3K9me3 observed in HADs. Western blot analysis found no expression changes in any HP1 isoform in Hat1 KO cells (Figure S3B-D), sugges1ng that HP1β is redistributed to HADs from other regions of the genome upon Hat1 KO. This may help explain the loss of HP1β in other LADs. Interes1ngly, the increases in H3K9me3 and HP1β in HADs seem to be independent of one another, as there is an insignificant degree of overlap between the sets of HADs that gain them and no correla1on between their respec1ve log2 fold-changes (Figure S3E). Together, these data suggest that Hat1 regulates the paTerns of repressive histone methyla1on in lamina-associated chroma1n, including the conversion of H3K9me2 to H3K9me3 and a concomitant increase in HP1b in HADs. Hat1 regulates the distribu.on of H3K9 -specific HMTs in LADs One mechanism by which Hat1 may influence the distribu1on of H3K9me3 and H3K9me2 throughout lamina-associated cons1tu1ve heterochroma1n is through regula1ng H3K9-specific .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 17 histone methyltransferases (HMTs). Hat1-dependent acetyla1on has been shown to control the stability of several proteins by regula1ng protein ubiquityla1on and proteosome-mediated degrada1on (73). However, Western blot analysis demonstrated that loss of Hat1 did not alter the abundance of the H3K9-specific HMTs Suv39h1, Suv39h2, G9a, or Setdb1 (Figure S4A-C). We next determined whether Hat1 regulates the genome-wide distribu1on of H3 K9-specific HMTs. We used CUT&RUN to profile Suv39h1, Suv39h2, G9a, and Setdb1 in the same HAT1 KO and WT iMEF cells lines that were used to profile the histone PTMs described above. In Hat1 WT cells, contrary to expecta1ons, we found that Suv39h1 did not coincide with the broad domains of H3K9me3 found in LADs. Rather, Suv39h1 localized primarily to discrete peaks outside of LADs, oben in euchroma1c regions, and was largely absent from most LADs (Figure 4A and 4B). In contrast, while Suv39h2 colocalized with Suv39h1 in a subset of peaks in euchroma1n, it was primarily found in broad domains that corresponded to domains of H3K9me3 in LADs (Figure 4A and 4B). Regression analysis showed that the abundance of Suv39h2 in LADs was closely correlated with the level of H3K9me3, while there is liTle correla1on between H3K9me3 and Suv39h1 in LADs (Figure 4C). Setdb1 also tended to be depleted in LADs, resembling the distribu1on of Suv39h1. Regression analysis indicated that there is no correla1on between H3K9me3 in LADs and the level of Setdb1 (Figure 4C). G9a was generally depleted in LADs, like Suv39h1 and Setdb1. However, regions of G9a accumula1on oben appeared inside LADs which showed enrichment inside the border and decreased toward the LAD interior (Figure 4A). This paTern was highly similar to that observed for H3K9me2. Indeed, we found that G9a was highly correlated with H3K9me2 inside LADs, but not correlated outside LADs (Figure S4D). Interes1ngly, the level of Suv39h2 across individual LADs oben appeared nega1vely correlated with that of G9a. Linear regression confirmed this nega1ve correla1on (Figure S4E), while the opposite trend was seen in inter-LAD regions. Not only was Suv39h2 generally lowest near the borders, where G9a was the highest, but peaks of G9a observed near the center of LADs oben corresponded with a dip in an otherwise uniform distribu1on of Suv39h2 (Figure 4D). These results suggest that, in iMEFs, Suv39h2 is primarily responsible for H3K9me3 in LADs and G9a is primarily responsible for H3K9me2 in LADs. Comparing HMT levels in Hat1 WT and KO cells, we found that Hat1 loss significantly increased Suv39h2 levels in HADs, while G9a levels were reduced. Changes in in the levels of Suv39h2 and G9a corresponded to changes in H3K9me3 and H3K9me2, respec1vely. This is demonstrated by the genome browser view shown in Figure 4D, which displays an ~9 Mb sec1on of chromosome 2. This region contains a large LAD with HADs located at each end. In the WT cells, there were high levels of H3K9me2 inside the borders of the LAD, in the regions designated as HADs. In Hat1 KO cells, the levels of both H3K9me2 and G9a drop significantly. These regions also showed corresponding increases in H3K9me3 and Suv39h2. There was liTle or no change in the levels of either Suv39h1 or Setdb1. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 18 Hat1-dependent changes in in H3K9-specific HMTs were not limited to HADs, though they were most pronounced there, on average (Figure 4E, Figure S4F). We profiled the localiza1on of Suv39h1, Suv39h2, G9a, and Setdb1 in Hat1 WT and Hat1 KO cells across all LADs. These profiles revealed changes across most LADs, as well as dis1nct characteris1cs of HADs. While the level of Suv39h1 was low across all LADs, it was higher in nhLADs rela1ve to HADs. Loss of Hat1 reduced the level of Suv39h1 across both nhLADs and HADs, similar to what was observed for H4K5/12ac and H3K9/27ac. The level of Suv39h2 is much higher in HADs than nhLADs. Hat1 KO increases Suv39h2 levels in both, but especially HADs. Conversely, G9a levels were higher in nhLADs than HADs, and decreased in both in Hat1 KO cells. The level of Setdb1 was higher in nhLADs than HADs but Hat1 had no effect on Setdb1 localiza1on. The Hat1-dependent changes in HMT localiza1on to LADs may be linked to the Hat1-dependent changes in histone methyla1on observed in these domains. Regression analyses show there was a strong correla1on between Hat1-dependent changes in G9a and H3K9me2 (Figure S5A). Similarly, there was a strong correla1on between changes in Suv39h2 and changes in H3K9me3 (Figure S5B). There was no correla1on between Hat1-dependent changes in Suv39h1 or Setdb1 and H3K9me3 (Figure S5C,D) In addi1on, the Hat1-dependent loss of G9a may be linked to the increase in Suv39h2. Regression analysis of the change in G9a and Suv39h2 showed a strongly nega1ve correla1on (Figure S5E). Together these results show that Hat1 influences the paTerns of H3K9 methyla1on in LADs by controlling the localiza1on of histone methyltransferases. chromHMM reveals that Hat1 restrains a shiS to more strongly heterochroma.c states ChromHMM is a Hidden Markov Model-based chroma1n classifica1on tool which groups regions of the genome into states based on the adjacent states and which features are enriched. To more clearly see how Hat1 alters the character of chroma1n across the genome, we ran chromHMM on our genomic maps of Lamin B1, HP1β, histone methyltransferases, and histone modifica1ons using 10 kb bins with a 10 state model (Figure 5A). The model was trained on the Hat1 WT and Hat1 KO data together to ensure the states were the same and the two condi1ons could be compared. The output states formed a rough con1nuum from euchroma1n enriched in histone acetyla1on (states 1 and 2) to cons1tu1ve heterochroma1n-enriched and lamina-associated (states 9 and 10). State 3 appeared to be an intermediate state that was similar to states 1 and 2 but was also enriched for H3K14ac and H3K9me2. States 5-7 were enriched to varying degrees in H3K27me3, H3K14ac, and H3K9me2, marks enriched inside LAD borders and across some LAD bodies. States 4 and 8 show very low enrichment of all features due to the nature of the algorithm used in chromHMM, which categorizes marks as simply present or absent in a given bin, placing many bins with infrequent enrichment of marks in apparently featureless states. The transi1ons .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 19 matrix showed that state 4 usually represented euchroma1c regions of low feature enrichment, while state 8 represented repressive regions with low enrichment of heterochroma1n features. Across the genome as a whole, Hat1 KO chroma1n showed a significant increase in the abundance of states 9 and 10, corresponding to cons1tu1ve heterochroma1n, with milder increases in euchroma1c states 2 and 3. There was a notable reduc1on in states 6 and 8, which represent a more moderate cons1tu1ve heterochroma1n signature (Figure 5B and 5C). WT LADs were predominantly enriched in states 6 – 9. The propor1on of LAD area occupied by states 6 – 10 remained similar in Hat1 KO cells, but there was a general transi1on towards higher states, with states 9 and 10, which are found in the interior of LADs, experiencing the greatest rela1ve growth at the expense of states 6 – 8, which predominate near LAD borders (Figure 5D). A similar shib was seen in HADs, which underwent a significant movement from states 7 – 9 to states 9 and 10, especially 9 (Figure 5E). These changes indicate that loss of Hat1

in a shib of exis1ng heterochroma1n states to states with greater enrichment of repressive cons1tu1ve heterochroma1n features. Hat1 colocalizes with Nuclear Lamina Proteins Our genomic analyses indicate that the most prominent effect of Hat1 is on nuclear lamina- associated domains of cons1tu1ve heterochroma1n. To determine whether HAT1 is directly linked to the nuclear lamina, we used APEX2-based proximity labeling to iden1fy proteins that localize near HAT1. APEX2 is an ascorbate peroxidase that generates a very short-lived bio1n- phenoxyl radical in the presence of bio1n phenol and H2O2 that can modify proteins in close proximity (Hung et al., 2016)(74,75). We transfected HEK293T cells with constructs that expressed either an APEX2-HAT1 fusion protein or APEX2 alone to control for non-specific bio1nyla1on. To confirm that the APEX2-Hat1 construct exhibited a paTern of localiza1on similar to that of endogenous HAT, we performed subcellular frac1ona1on. Upon staining with an an1-HAT1 an1body, we observed that the APEX2-HAT1 fusion was localized to both the cytoplasm and the nucleus, similar to the localiza1on of endogenous HAT1. Importantly, the APEX2-HAT1 fusion protein was expressed at a level comparable to endogenous HAT1 in both the nucleus and the cytoplasm (Supplementary Figure S6). Following incuba1on of cells with bio1n phenol and H2O2, extracts were generated, bio1nylated proteins were isolated by streptavidin purifica1on and iden1fied by mass spectrometry. A total of 29 proteins were found to be significantly bio1nylated by APEX2-HAT1 aber filtering out the proteins that were significantly bio1nylated in control cells (untransfected cells and cells expressing APEX2 only). A volcano plot of significantly bio1nylated proteins included HAT1 and histone H4, suppor1ng the validity of this approach (Figure 6A). PCNA was also significantly bio1nylated, consistent with recent reports demonstra1ng that HAT1 transiently localizes to newly replicated DNA (Agudelo Garcia et al., 2017; Agudelo Garcia et al., .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 20 2020a; Nagarajan et al., 2013). Importantly, LBR (lamin B receptor) and EMD (emerin), a protein that links the nuclear lamina to the nuclear membrane, were also found to be significantly bio1nylated by APEX2-HAT1. The iden1fica1on of LBR in proximity to Hat1 I consistent with the reent iden1fica1on of LBR as a direct substrate of Hat1(76). These results suggest that a popula1on of HAT1 resides in proximity to the nuclear lamina. We next used a proximity liga1on assay (PLA) to directly visualize the associa1on of HAT1 with the nuclear lamina in cells. PLAs determine whether two molecules reside close to each other in the cell by employing two species-specific secondary an1bodies that are fused to oligonucleo1des. If the secondary an1bodies recognize primary an1bodies that are in close proximity, the oligonucleo1des can both bind to a nicked circular DNA, crea1ng a template for rolling circle replica1on. This amplifies sequences that can be bound by a fluorescent probe and visualized. As a posi1ve control, we performed a PLA using two different an1bodies that each recognize lamina B1. As seen in Figure 6B, a posi1ve PLA signal is detected in both HAT1 WT and HAT1 KO iMEFs. We then combined one of the lamin B1 an1bodies with an an1body recognizing HAT1. PLA signals are readily detected in HAT1 WT iMEFs but are lost in HAT1 KO iMEFs, confirming that HAT1 localized to the nuclear lamina (Figure 6C).

The reestablishment of heterochroma1n aber S-phase is a gradual process which takes the beTer part of a cell cycle (27,51,77). In recent years, studies have begun to reveal the mechanisms involved. Suv39h1/2 and HP1 mutually promote one-another’s reten1on at heterochroma1n, leading to transcrip1onal silencing, deposi1on of H3K9me3, and the recruitment of chroma1n remodelers, HDACs, histone methyltransferases, and more (78-83). Many factors associated with the replica1on fork are essen1al for heterochroma1n regula1on because of their role in recycling parental histones(84-86). Chroma1n assembly factor 1 (CAF-1) also plays a crucial role by deposi1ng new histones to prevent nucleosome deple1on across the genome(87,88). However, an ac1ve role for new histone modifica1ons in controlling chroma1n re-establishment has not been recognized. Using custom Hat1 KO cell lines which lack nascent H4K5/12ac, we show that this loss leads to complex changes in lamina-associated chroma1n, with a net shib towards more repressive states. Lack of Hat1 causes an enrichment of Suv39h2 and corresponding deple1on of G9a in LADs, which leads to increased H3K9me3 and loss of H3K9me2. Interes1ngly, not all LADs respond to Hat1 loss in iden1cal fashion. While almost all experience some reduc1on in histone acetyla1on, the shib to a stronger cons1tu1ve heterochroma1n state is par1cularly seen in a subset of LADs, many of which were previously iden1fied as HADs. While LADs in iMEFs can be defined by enrichment of either H3K9me3, H3K9me2, or H3K27me3, HADs fall almost en1rely into the first of these categories. HADs are enriched in H3K9me3 and .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 21 HP1β, depleted in H3K27me3, and above average size for LADs. It appears that the increase in Suv39h2 is taking place largely in regions where it was already enriched, and we are observing enhanced, not de novo recruitment. However, H3K9me3 abundance is not sufficient to explain why only this subset of LADs gains Suv39h2, as there are LADs with just as high or higher Suv39h2 levels which do not gain it upon Hat1 KO. We also see inconsistency in whether or not LADs gain HP1 upon Hat1 loss. We only examined the HP1β isoform. It would be interes1ng to see whether the other two HP1 isoforms increase in heterochroma1n, par1cularly HP1α, which, like HP1β, is enriched in chromocenters (89). We were unable to determine a single structural feature which would explain the greater sensi1vity of some LADs to Hat1 loss. Numerous factors, including the histone PTM signature, DNA methyla1on, 3D chroma1n interac1ons, lamina-associa1on, transcrip1on, or sequence-specific DNA binding proteins, may dis1nguish them from other LADs, and it is likely that there are mul1ple contribu1ng factors. Suv39h1 and Su39h2 share the same general structure apart from an 81 amino acid N-terminal basic domain which is present only in Suv39h2 (90,91). As a result, they are oben treated as interchangeable, though numerous studies have revealed differences in their substrate specificity, strength of localiza1on to heterochroma1n, and silencing capacity (92-94). One study found that Suv39h2 is primarily responsible for the deposi1on of H3K9me3 at cons1tu1ve heterochroma1n and localizes to chromocenters more independently than Suv39h1 does, while Suv39h1 contributes more to transcrip1onal silencing(94). Our finding that only Suv39h2 is enriched in cons1tu1ve heterochroma1n while Suv39h1 is more enriched in genic regions is mostly consistent with this. However, immunofluorescence has suggested both isoforms to be enriched in chromocenters, without significant localiza1on to the nuclear periphery. This may point to differences in Suv39h1 and Suv39h2 localiza1on between different cell types or cell lines. The unique N-terminal domain of Suv39h2 can bind RNA. Major satellite repeat transcripts help recruit it to pericentric regions but may also reduce its ac1vity in some contexts (91,95). It would be interes1ng to test whether non-coding RNA transcripts contribute to Suv39h2 recruitment in our MEFs, though increased transcrip1on upon loss of H4K5/12ac would be unexpected. The observed loss of acetyla1on in LADs upon Hat1 KO is interes1ng since LADs are generally perceived as lacking histone acetyla1on. There are at least two ways to interpret this finding which are not mutually exclusive: (1) LADs maintain some basal level of histone acetyla1on throughout the cell cycle, possibly in order to preserve plas1city and the poten1al for ac1va1on, and thus are regulated by opposing pathways that preserve heterochroma1n while ensuring that they are not as heterochroma1c as they could be. (2) A specific subpopula1on of cells possesses histone acetyla1on in LADs but cannot maintain it in the absence of Hat1. The second interpreta1on is consistent with S/early G2 phase cells being enriched in acetylated, newly synthesized histones. We have previously shown that nascent H4K5/12ac is eliminated in Hat1 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 22 KO cells and that Hat1 loss indirectly leads to deple1on of H3K9 and K27 acetyla1on on nascent chroma1n (58). While H4K5/12ac acetyla1on is probably reduced genome-wide in these cells, the rela1ve effect is probably much more pronounced in heterochroma1n, where parental histones are rarely acetylated, than in euchroma1n, where acetyla1on is already enriched on many parental histones. There are several ways in which acetyla1on of nascent chroma1n might regulate heterochroma1n matura1on. H4K5/12ac can recruit bromodomain proteins such as ATAD2, which competes with HDACs 1 and 2 and influences HP1 recruitment, or BRPF3, which complexes with the H3K14 acetyltransferase KAT7 (96-99). We previously found that without Hat1, nascent chroma1n was depleted in Brd3 and Brg1, both readers of acetyla1on on H4 K5 and/or K12 (61). It is also possible that H4K5/12ac affects the produc1on of nascent chroma1n transcripts, which helps recruit regulatory proteins and promote chroma1n matura1on(100- 102). Hat1 or H4K5/12ac on new histones may even directly influence the recruitment or cataly1c ac1vity of H3K9 methyltransferases. As suggested above, Hat1 may also regulate the presence of other acetyla1on marks on nascent chroma1n besides H4K5/12ac, which could have numerous downstream effects. Acetyla1on of H3K14 or H3K18 is found on over 20% of new histone H3 in human cells, and it is unknown whether they are deposited across the genome or biased toward specific regions(52). Ubiquityla1on of H3K18 and H3K14, especially the laTer, can promote H3K9me3 by Suv39h1/2, and their acetyla1on would block this ubiquityla1on (46,47,103,104). In this context, it is interes1ng that the paTern and Hat1- sensi1vity of H3K14ac is markedly different between HADs and nhLADs. H3K9ac could directly slow the spread of heterochroma1n by reducing the amount of K9 available to be methylated. Since H3K9me3 must exceed a cri1cal density in order to be propagated by the read-write mechanism, it is possible that small differences in H3K9ac could drama1cally affect H3K9 spreading, especially near the threshold level (44). Cons1tu1ve heterochroma1n spreads via a posi1ve feedback loop in which H3K9me3 recruits repressors such as HP1 proteins or Suv39h1/2 which promote the deposi1on of more H3K9me3 on nearby nucleosomes. The increased H3K9me3 observed in Hat1 KO HADs raises the ques1on of why cons1tu1ve heterochroma1n spreading into neighboring euchroma1n is not observed. LAD boundaries remain remarkably consistent in the absence of Hat1, even those which gain H3K9me3 and Suv39h2. A poten1al explana1on is the increased H4K5ac observed just outside HAD/nhLAD borders in Hat1 KO cells. This suggests that a protec1ve mechanism exists in cells that senses the chroma1n state of LAD borders and can respond to increased H3K9me3 by the recruitment of other HATs that prevent heterochroma1n spreading. It is uncertain what frac1on of nucleosomes are trimethylated in a par1cular region of the genome. Our results suggest that nucleosomes in HADs are not saturated with H3K9me3, given the increase observed upon Hat1 KO. It is interes1ng to note, however, that LADs which gain .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 23 H3K9me3 in Hat1 KO cells tend have much lower WT levels of H3K9me3 on average than those in which H3K9me3 is unchanged or lost (Figure S7). This suggests that LADs which are closest to satura1ng levels of H3K9me3 are less likely to gain more upon Hat1 loss. Alterna1vely, we may be observing increased H3K9me3 in a subset of cells which had lower levels than the rest of the popula1on. Despite the numerous chroma1n changes caused by loss of Hat1, the changes in gene expression are rela1vely few (Popova et al 2021), probably because most of the chroma1n changes take place in very gene poor regions. This seems to conflict with the vital importance of Hat1 for mammalian development. We hypothesize that Hat1 must moderate heterochroma1n re-establishment to allow binding of lineage-specific transcrip1on factors and facilitate proper differen1a1on during animal development. This is supported by our recent findings that Hat1 loss interferes with differen1a1on of intes1nal stem cells in vivo and in vitro, and produces enhanced domains of H3K9me3 analogous to those described in this study (Nagarajan et al BIORXIV/2026/712164). While many interes1ng ques1ons remain, the results of this study supply an important advance in our understanding of Hat1’s role in epigenome regula1on. Hat1 contributes to the regula1on of H3K9 methyltransferase recruitment and regulates the balance of ac1ve and repressive features in heterochroma1n, making it a vital regulator of chroma1n inheritance. Data Availability ATAC-seq data, HAD loca1ons and LAD loca1ons used for our analysis were previously published and are deposited in Gene Expression Omnibus GSE178592 and GSE281928 (64,65). All data generated in this study has been deposited in the NCBI Gene Expression Omnibus under accession numbers GSE325188 (CUT&Tag), GSE325189 (CUT&RUN), GSE325190 (ChIP-seq), and GSE325191 (RNA- seq). Funding This work was supported by grant R01 GM144601 to MRP. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 24

1. Escobar, T.M., Loyola, A. and Reinberg, D. (2021) Parental nucleosome segrega)on and the inheritance of cellular iden)ty. Nature Reviews Gene/cs, 22, 379-392. 2. Kloc, A., Zara)egui, M., Nora, E. and Mar)enssen, R. (2008) RNA Interference Guides Histone Modifica)on during the S Phase of Chromosomal Replica)on. Current Biology, 18, 490-495. 3. Grewal, S.I.S. (2023) The molecular basis of heterochroma)n assembly and epigene)c inheritance. Molecular Cell, 83, 1767-1785. 4. Ramachandran, S. and Henikoff, S. (2016) Transcrip)onal Regulators Compete with Nucleosomes Post-replica)on. Cell, 165, 580-592. 5. Bell, O., Burton, A., Dean, C., Gasser, S.M. and Torres-Padilla, M.-E. (2023) Heterochroma)n defini)on and func)on. Nat Rev Mol Cell Biol, 24, 691-694. 6. Nothof, S.A., Magdinier, F. and Van-Gils, J. (2022) Chroma)n Structure and Dynamics: Focus on Neuronal Differen)a)on and Pathological Implica)on. Genes, 13, 639. 7. Kim, J.J. and Kingston, R.E. (2022) Context-specific Polycomb mechanisms in development. Nat Rev Genet, 23, 680-695. 8. Blackledge, N.P . and Klose, R.J. (2021) The molecular principles of gene regula)on by Polycomb repressive complexes. Nat Rev Mol Cell Biol, 22, 815-833. 9. Bhat, M.Y . and Liu, X. (2026) Emerging structural insights into PRC2 func)on in development and disease. Trends Biochem Sci, 51, 172-186. 10. Janssen, A., Colmenares, S.U. and Karpen, G.H. (2018) Heterochroma)n: Guardian of the Genome. Annu Rev Cell Dev Biol, 34, 265-288. 11. Sanulli, S., Trnka, M.J., Dharmarajan, V., Tibble, R.W., Pascal, B.D., Burlingame, A.L., Griffin, P .R., Gross, J.D. and Narlikar, G.J. (2019) HP1 reshapes nucleosome core to promote phase separa)on of heterochroma)n. Nature, 575, 390-394. 12. Machida, S., Takizawa, Y ., Ishimaru, M., Sugita, Y ., Sekine, S., Nakayama, J.I., Wolf, M. and Kurumizaka, H. (2018) Structural Basis of Heterochroma)n Forma)on by Human HP1. Mol Cell, 69, 385-397 e388. 13. Briand, N. and Collas, P . (2020) Lamina-associated domains: peripheral makers and internal affairs. Genome Biol, 21, 85. 14. Kiseleva, A.A., Cheng, Y .C., Smith, C.L., Katz, R.A. and Poleshko, A. (2023) PRR14 organizes H3K9me3-modified heterochroma)n at the nuclear lamina. Nucleus, 14, 2165602. 15. Ye, Q., Callebaut, I., Pezhman, A., Courvalin, J.C. and Worman, H.J. (1997) Domain-specific interac)ons of human HP1-type chromodomain proteins and inner nuclear membrane protein LBR. J Biol Chem, 272, 14983-14989. 16. Harr, J.C., Luperchio, T.R., Wong, X., Cohen, E., Wheelan, S.J. and Reddy, K.L. (2015) Directed targe)ng of chroma)n to the nuclear lamina is mediated by chroma)n state and A-type lamins. Journal of Cell Biology, 208, 33-52. 17. Peric-Hupkes, D., Meuleman, W., Pagie, L., Bruggeman, S.W., Solovei, I., Brugman, W., Graf, S., Flicek, P ., Kerkhoven, R.M., van Lohuizen, M. et al. (2010) Molecular maps of the reorganiza)on of genome-nuclear lamina interac)ons during differen)a)on. Mol Cell, 38, 603-613. 18. Madsen-Osterbye, J., Abdelhalim, M., Baudement, M.O. and Collas, P . (2022) Local euchroma)n enrichment in lamina-associated domains an)cipates their reposi)oning in the adipogenic lineage. Genome Biol, 23, 91. 19. Zheng, X., Hu, J., Yue, S., Kris)ani, L., Kim, M., Sauria, M., Taylor, J., Kim, Y . and Zheng, Y . (2018) Lamins Organize the Global Three-Dimensional Genome from the Nuclear Periphery. Mol Cell, 71, 802-815 e807. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 25 20. Moir, R.D., Spann, T.P ., Herrmann, H. and Goldman, R.D. (2000) Disrup)on of nuclear lamin organiza)on blocks the elonga)on phase of DNA replica)on. J Cell Biol, 149, 1179-1192. 21. Solovei, I., Wang, A.S., Thanisch, K., Schmidt, C.S., Krebs, S., Zwerger, M., Cohen, T.V., Devys, D., Foisner, R., Peichl, L. et al. (2013) LBR and lamin A/C sequen)ally tether peripheral heterochroma)n and inversely regulate differen)a)on. Cell, 152, 584-598. 22. Ulianov, S.V., Doronin, S.A., Khrameeva, E.E., Kos, P .I., Luzhin, A.V., Starikov, S.S., Galitsyna, A.A., Nenasheva, V.V., Ilyin, A.A., Flyamer, I.M. et al. (2019) Nuclear lamina integrity is required for proper spa)al organiza)on of chroma)n in Drosophila. Nat Commun, 10, 1176. 23. Wang, Y., Elsherbiny, A., Kessler, L., Cordero, J., Shi, H., Serke, H., Lityagina, O., Trogisch, F.A., Mohammadi, M.M., El-Bakrawy, I. et al. (2022) Lamin A/C-dependent chroma)n architecture safeguards naive pluripotency to prevent aberrant cardiovascular cell fate and func)on. Nat Commun, 13, 6663. 24. Kohwi, M., Lupton, J.R., Lai, S.L., Miller, M.R. and Doe, C.Q. (2013) Developmentally regulated subnuclear genome reorganiza)on restricts neural progenitor competence in Drosophila. Cell, 152, 97-108. 25. Poleshko, A., Shah, P.P., Gupta, M., Babu, A., Morley, M.P., Manderfield, L.J., Imovits, J.L., Calderon, D., Aghajanian, H., Sierra-Pagan, J.E. et al. (2017) Genome-Nuclear Lamina Interac)ons Regulate Cardiac Stem Cell Lineage Restric)on. Cell, 171, 573-587 e514. 26. Gonzalez-Sandoval, A., Towbin, B.D., Kalck, V., Cabianca, D.S., Gaidatzis, D., Hauer, M.H., Geng, L., Wang, L., Yang, T., Wang, X. et al. (2015) Perinuclear Anchoring of H3K9-Methylated Chroma)n Stabilizes Induced Cell Fate in C. elegans Embryos. Cell, 163, 1333-1347. 27. Reverón-Gómez, N., González-Aguilera, C., Stewart-Morgan, K.R., Petryk, N., Flury, V., Graziano, S., Johansen, J.V., Jakobsen, J.S., Alabert, C. and Groth, A. (2018) Accurate Recycling of Parental Histones Reproduces the Histone Modifica)on Landscape during DNA Replica)on. Molecular Cell, 72, 239-249.e235. 28. Shibahara, K. and S)llman, B. (1999) Replica)on-dependent marking of DNA by PCNA facilitates CAF-1-coupled inheritance of chroma)n. Cell, 96, 575-585. 29. Sauer, P .V., Timm, J., Liu, D., Sitbon, D., Boeri-Erba, E., Velours, C., Mücke, N., Langowski, J., Ochsenbein, F., Almouzni, G. et al. (2017) Insights into the molecular architecture and histone H3-H4 deposi)on mechanism of yeast Chroma)n assembly factor 1. eLife, 6, e23474. 30. Padeken, J., Methot, S.P. and Gasser, S.M. (2022) Establishment of H3K9-methylated heterochroma)n and its func)ons in )ssue differen)a)on and maintenance. Nat Rev Mol Cell Biol. 31. Montavon, T., Shukeir, N., Erikson, G., Engist, B., Onishi-Seebacher, M., Ryan, D., Musa, Y ., Mikler, G., Meyer, A.G., Genoud, C. et al. (2021) Complete loss of H3K9 methyla)on dissolves mouse heterochroma)n organiza)on. Nature Communica/ons, 12, 4359. 32. Rice, J.C., Briggs, S.D., Ueberheide, B., Barber, C.M., Shabanowitz, J., Hunt, D.F., Shinkai, Y . and Allis, C.D. (2003) Histone Methyltransferases Direct Different Degrees of Methyla)on to Define Dis)nct Chroma)n Domains. Molecular Cell, 12, 1591-1598. 33. Loyola, A., Tagami, H., Bonaldi, T., Roche, D., Quivy, J.P ., Imhof, A., Nakatani, Y ., Dent, S.Y .R. and Almouzni, G. (2009) The HP1alpha-CAF1-SetDB1-containing complex provides H3K9me1 for Suv39-mediated K9me3 in pericentric heterochroma)n. EMBO reports, 10, 769-775. 34. Elsässer, S.J., Noh, K.-M., Diaz, N., Allis, C.D. and Banaszynski, L.A. (2015) Histone H3.3 is required for endogenous retroviral element silencing in embryonic stem cells. Nature, 522, 240-244. 35. Matsui, T., Leung, D., Miyashita, H., Maksakova, I.A., Miyachi, H., Kimura, H., Tachibana, M., Lorincz, M.C. and Shinkai, Y . (2010) Proviral silencing in embryonic stem cells requires the histone methyltransferase ESET. Nature, 464, 927-931. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 26 36. Karimi, M.M., Goyal, P ., Maksakova, I.A., Bilenky, M., Leung, D., Tang, J.X., Shinkai, Y., Mager, D.L., Jones, S., Hirst, M. et al. (2011) DNA methyla)on and SETDB1/H3K9me3 regulate predominantly dis)nct sets of genes, retroelements, and chimeric transcripts in mESCs. Cell Stem Cell, 8, 676- 687. 37. Al-Sady, B., Madhani, H.D. and Narlikar, G.J. (2013) Division of labor between the chromodomains of HP1 and Suv39 methylase enables coordina)on of heterochroma)n spread. Molecular Cell, 51, 80-91. 38. Audergon, P .N.C.B., Catania, S., Kagansky, A., Tong, P ., Shukla, M., Pidoux, A.L. and Allshire, R.C. (2015) Restricted epigene)c inheritance of H3K9 methyla)on. Science, 348, 132-135. 39. Ragunathan, K., Jih, G. and Moazed, D. (2015) Epigene)c inheritance uncoupled from sequence- specific recruitment. Science, 348, 1258699. 40. Müller, M.M., Fierz, B., Bikova, L., Liszczak, G. and Muir, T.W. (2016) A two-state ac)va)on mechanism controls the histone methyltransferase Suv39h1. Nat Chem Biol, 12, 188-193. 41. Hansen, K.H., Bracken, A.P ., Pasini, D., Dietrich, N., Gehani, S.S., Monrad, A., Rappsilber, J., Lerdrup, M. and Helin, K. (2008) A model for transmission of the H3K27me3 epigene)c mark. Nat Cell Biol, 10, 1291-1300. 42. Moussa, H.F., Bsteh, D., Yelagandula, R., Pribitzer, C., Stecher, K., Bartalska, K., Micher, L., Wang, J., Zepeda-Mar)nez, J.A., Elling, U. et al. (2019) Canonical PRC1 controls sequence-independent propaga)on of Polycomb-mediated gene silencing. Nature Communica/ons, 10, 1931. 43. Fukuda, K., Shimura, C. and Shinkai, Y . (2025) H3K27me3 and the PRC1-H2AK119ub pathway coopera)vely maintain heterochroma)n and transcrip)onal silencing aser the loss of H3K9 methyla)on. Epigene/cs Chroma/n, 18, 26. 44. Cuker DiPiazza, A.R., Taneja, N., Dhakshnamoorthy, J., Wheeler, D., Holla, S. and Grewal, S.I.S. (2021) Spreading and epigene)c inheritance of heterochroma)n require a cri)cal density of histone H3 lysine 9 tri-methyla)on. Proceedings of the Na/onal Academy of Sciences, 118, e2100699118. 45. Owen, J.A., Osmanović, D. and Mirny, L. (2023) Design principles of 3D epigene)c memory systems. Science, 382, eadg3053. 46. S)rpe, A., Guidor, N., Northall, S.J., Kilic, S., Hainard, A., Vadas, O., Fierz, B. and Schalch, T. (2021) SUV39 SET domains mediate crosstalk of heterochroma)c histone marks. eLife, 10, e62682. 47. Oya, E., Nakagawa, R., Yoshimura, Y ., Tanaka, M., Nishibuchi, G., Machida, S., Shirai, A., Ekwall, K., Kurumizaka, H., Tagami, H. et al. (2019) H3K14 ubiquityla)on promotes H3K9 methyla)on for heterochroma)n assembly. EMBO reports, 20, e48111. 48. Liu, C., Yu, J., Song, A., Wang, M., Hu, J., Chen, P ., Zhao, J. and Li, G. (2023) Histone H1 facilitates restora)on of H3K27me3 during DNA replica)on by chroma)n compac)on. Nature Communica/ons, 14, 4081. 49. Willcockson, M.A., Healton, S.E., Weiss, C.N., Bartholdy, B.A., Botbol, Y ., Mishra, L.N., Sidhwani, D.S., Wilson, T.J., Pinto, H.B., Maron, M.I. et al. (2021) H1 histones control the epigene)c landscape by local chroma)n compac)on. Nature, 589, 293-298. 50. Alabert, C., Bukowski-Wills, J.-C., Lee, S.-B., Kustatscher, G., Nakamura, K., de Lima Alves, F., Menard, P ., Mejlvang, J., Rappsilber, J. and Groth, A. (2014) Nascent chroma)n capture proteomics determines chroma)n dynamics during DNA replica)on and iden)fies unknown fork components. Nat Cell Biol, 16, 281-291. 51. Zee, B.M., Brikon, L.-M.P ., Wolle, D., Haberman, D.M. and Garcia, B.A. (2012) Origins and Forma)on of Histone Methyla)on across the Human Cell Cycle. Molecular and Cellular Biology, 32, 2503-2514. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 27 52. Jasencakova, Z., Scharf, A.N.D., Ask, K., Corpet, A., Imhof, A., Almouzni, G. and Groth, A. (2010) Replica)on stress interferes with histone recycling and predeposi)on marking of new histones. Molecular Cell, 37, 736-743. 53. Parthun, M.R., Widom, J. and Gokschling, D.E. (1996) The major cytoplasmic histone acetyltransferase in yeast: links to chroma)n replica)on and histone metabolism. Cell, 87, 85-94. 54. Verreault, A., Kaufman, P.D., Kobayashi, R. and S)llman, B. (1998) Nucleosomal DNA regulates the core-histone-binding subunit of the human Hat1 acetyltransferase. Curr Biol, 8, 96-108. 55. Gruber, J.J., Geller, B., Lipchik, A.M., Chen, J., Salahudeen, A.A., Ram, A.N., Ford, J.M., Kuo, C.J. and Snyder, M.P . (2019) HAT1 Coordinates Histone Produc)on and Acetyla)on via H4 Promoter Binding. Molecular Cell, 75, 711-724.e715. 56. Sobel, R.E., Cook, R.G., Perry, C.A., Annunziato, A.T. and Allis, C.D. (1995) Conserva)on of deposi)on-related acetyla)on sites in newly synthesized histones H3 and H4. Proceedings of the Na/onal Academy of Sciences, 92, 1237-1241. 57. Varga, J., Korbai, S., Neller, A., Zsindely, N. and Bodai, L. (2019) Hat1 acetylates histone H4 and modulates the transcrip)onal program in Drosophila embryogenesis. Scien/fic Reports, 9, 17973. 58. Nagarajan, P ., Ge, Z., Sirbu, B., Doughty, C., Agudelo Garcia, P .A., Schlederer, M., Annunziato, A.T., Cortez, D., Kenner, L. and Parthun, M.R. (2013) Histone Acetyl Transferase 1 Is Essen)al for Mammalian Development, Genome Stability, and the Processing of Newly Synthesized Histones H3 and H4. PLOS Gene/cs, 9, e1003518. 59. Ai, X. and Parthun, M.R. (2004) The nuclear Hat1p/Hat2p complex: a molecular link between type B histone acetyltransferases and chroma)n assembly. Molecular Cell, 14, 195-205. 60. Barman, H.K., Takami, Y ., Nishijima, H., Shibahara, K.-i., Sanematsu, F. and Nakayama, T. (2008) Histone acetyltransferase-1 regulates integrity of cytosolic histone H3–H4 containing complex. Biochemical and Biophysical Research Communica/ons, 373, 624-630. 61. Agudelo Garcia, P .A., Hoover, M.E., Zhang, P ., Nagarajan, P ., Freitas, M.A. and Parthun, M.R. (2017) Iden)fica)on of mul)ple roles for histone acetyltransferase 1 in replica)on-coupled chroma)n assembly. Nucleic Acids Research, 45, 9319-9335. 62. Taddei, A., Roche, D., Sibarita, J.B., Turner, B.M. and Almouzni, G. (1999) Duplica)on and maintenance of heterochroma)n domains. The Journal of Cell Biology, 147, 1153-1166. 63. Sirbu, B.M., Couch, F.B., Feigerle, J.T., Bhaskara, S., Hiebert, S.W. and Cortez, D. (2011) Analysis of protein dynamics at ac)ve, stalled, and collapsed replica)on forks. Genes Dev., 25, 1320-1327. 64. Popova, L.V., Nagarajan, P., Lovejoy, C.M., Sunkel, B.D., Gardner, M.L., Wang, M., Freitas, M.A., Stanton, B.Z. and Parthun, M.R. (2021) Epigene)c regula)on of nuclear lamina-associated heterochroma)n by HAT1 and the acetyla)on of newly synthesized histones. 49, 12136-12151. 65. Mar)n, C.J., Oser, E.A., Nagarajan, P ., Popova, L.V., Sunkel, B.D., Stanton, B.Z. and Parthun, M.R. (2025) Dis)nct classes of lamina-associated domains are defined by differen)al pakerns of repressive histone methyla)on. Genome Res, 35, 1959-1974. 66. Kaya-Okur, H.S., Janssens, D.H., Henikoff, J.G., Ahmad, K. and Henikoff, S. (2020) Efficient low-cost chroma)n profiling with CUT&Tag. Nat Protoc, 15, 3264-3283. 67. Nagarajan, P ., Agudelo Garcia, P .A., Iyer, C.C., Popova, L.V., Arnold, W.D. and Parthun, M.R. (2019) Early-onset aging and mitochondrial defects associated with loss of histone acetyltransferase 1 (Hat1). Aging Cell, 18, e12992. 68. Chambers, M.C., Maclean, B., Burke, R., Amodei, D., Ruderman, D.L., Neumann, S., Gako, L., Fischer, B., Prak, B., Egertson, J. et al. (2012) A cross-plavorm toolkit for mass spectrometry and proteomics. Nat Biotechnol, 30, 918-920. 69. Rost, H.L., Sachsenberg, T., Aiche, S., Bielow, C., Weisser, H., Aicheler, F., Andreor, S., Ehrlich, H.C., Gutenbrunner, P ., Kenar, E. et al. (2016) OpenMS: a flexible open-source sosware plavorm for mass spectrometry data analysis. Nat Methods, 13, 741-748. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 28 70. Gardner, M.L. and Freitas, M.A. (2021) Mul)ple Imputa)on Approaches Applied to the Missing Value Problem in Bokom-Up Proteomics. Interna/onal journal of molecular sciences, 22. 71. Lazarchuk, P ., Roy, S., Schlacher, K. and Sidorova, J. (2019) Detec)on and Quan)ta)on of Acetylated Histones on Replica)ng DNA Using In Situ Proximity Liga)on Assay and Click-It Chemistry. Methods Mol Biol, 1983, 29-45. 72. Wang, Z., Zang, C., Rosenfeld, J.A., Schones, D.E., Barski, A., Cuddapah, S., Cui, K., Roh, T.-Y. , Pe n g , W., Zhang , M.Q. et al. (2008) Combinatorial pakerns of histone acetyla)ons and methyla)ons in the human genome. Nature Gene/cs, 40, 897-903. 73. Yuan, Y., Miao, Y., Qian, L., Zhang, Y., Liu, C., Liu, J., Zuo, Y., Feng, Q., Guo, T., Zhang, L. et al. (2020) Targe)ng UBE4A Revives Viperin Protein in Epithelium to Enhance Host An)viral Defense. Mol Cell, 77, 734-747 e737. 74. Lam, S.S., Martell, J.D., Kamer, K.J., Deerinck, T.J., Ellisman, M.H., Mootha, V.K. and Ting, A.Y . (2015) Directed evolu)on of APEX2 for electron microscopy and proximity labeling. Nat Methods, 12, 51-54. 75. Huang, M.S., Lin, W.C., Chang, J.H., Cheng, C.H., Wang, H.Y . and Mou, K.Y . (2019) The cysteine- free single mutant C32S of APEX2 is a highly expressed and ac)ve fusion tag for proximity labeling applica)ons. Protein Sci, 28, 1703-1712. 76. Song, J., Ngo, L., Bell, K. and Zheng, Y .G. (2022) Chemoproteomic Profiling of Protein Substrates of a Major Lysine Acetyltransferase in the Na)ve Cellular Context. ACS Chem Biol, 17, 1092-1102. 77. Xu, M., Wang, W., Chen, S. and Zhu, B. (2012) A model for mito)c inheritance of histone lysine methyla)on. EMBO Reports, 13, 60-67. 78. Eskeland, R., Eberharter, A. and Imhof, A. (2007) HP1 Binding to Chroma)n Methylated at H3K9 Is Enhanced by Auxiliary Factors. Molecular and Cellular Biology, 27, 453-465. 79. Cheu)n, T., McNairn, A.J., Jenuwein, T., Gilbert, D.M., Singh, P .B. and Misteli, T. (2003) Maintenance of stable heterochroma)n domains by dynamic HP1 binding. Science, 299, 721- 725. 80. Haldar, S., Saini, A., Nanda, J.S., Saini, S. and Singh, J. (2011) Role of Swi6/HP1 self-associa)on- mediated recruitment of Clr4/Suv39 in establishment and maintenance of heterochroma)n in fission yeast. J Biol Chem, 286, 9308-9320. 81. Fischer, T., Cui, B., Dhakshnamoorthy, J., Zhou, M., Rubin, C., Zofall, M., Veenstra, T.D. and Grewal, S.I. (2009) Diverse roles of HP1 proteins in heterochroma)n assembly and func)ons in fission yeast. Proc Natl Acad Sci U S A, 106, 8998-9003. 82. Leopold, K., S)rpe, A. and Schalch, T. (2019) Transcrip)onal gene silencing requires dedicated interac)on between HP1 protein Chp2 and chroma)n remodeler Mit1. Genes Dev, 33, 565-577. 83. Nakao, M., Sato, Y ., Aizawa, A. and Kimura, H. (2024) Mode of SUV420H2 heterochroma)n localiza)on through mul)ple HP1 binding mo)fs in the heterochroma)c targe)ng module. Genes Cells, 29, 361-379. 84. Yu, J., Zhang, Y., Fang, Y., Paulo, J.A., Yaghoubi, D., Hua, X., Shipkovenska, G., Toda, T., Zhang, Z., Gygi, S.P . et al. (2024) A replisome-associated histone H3-H4 chaperone required for epigene)c inheritance. Cell, S0092867424007669. 85. Bellelli, R., Belan, O., Pye, V.E., Clement, C., Maslen, S.L., Skehel, J.M., Cherepanov, P ., Almouzni, G. and Boulton, S.J. (2018) POLE3-POLE4 Is a Histone H3-H4 Chaperone that Maintains Chroma)n Integrity during DNA Replica)on. Molecular Cell, 72, 112-126.e115. 86. Toda, T., Fang, Y., Shan, C.M., Hua, X., Kim, J.K., Tang, L.C., Jovanovic, M., Tong, L., Qiao, F., Zhang, Z. et al. (2024) Mrc1 regulates parental histone segrega)on and heterochroma)n inheritance. Mol Cell, 84, 3223-3236 e3224. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 29 87. Cheloufi, S., Elling, U., Hopfgartner, B., Jung, Y .L., Murn, J., Ninova, M., Hubmann, M., Badeaux, A.I., Euong Ang, C., Tenen, D. et al. (2015) The histone chaperone CAF-1 safeguards soma)c cell iden)ty. Nature, 528, 218-224. 88. Hatanaka, Y. , I n o u e , K . , O i ka wa , M . , Ka m i m u ra , S . , O g o n u k i , N . , Ko d a m a , E . N . , O h ka wa , Y. , Tsukada, Y.-i. and Ogura, A. (2015) Histone chaperone CAF-1 mediates repressive histone modifica)ons to protect preimplanta)on mouse embryos from endogenous retrotransposons. Proc. Natl. Acad. Sci. U. S. A., 112, 14641-14646. 89. Minc, E., Allory, Y ., Worman, H.J., Courvalin, J.C. and Buendia, B. (1999) Localiza)on and phosphoryla)on of HP1 proteins during the cell cycle in mammalian cells. Chromosoma, 108, 220-234. 90. O'Carroll, D., Scherthan, H., Peters, A.H., Opravil, S., Haynes, A.R., Laible, G., Rea, S., Schmid, M., Lebersorger, A., Jerratsch, M. et al. (2000) Isola)on and characteriza)on of Suv39h2, a second histone H3 methyltransferase gene that displays tes)s-specific expression. Mol Cell Biol, 20, 9423-9433. 91. Velazquez Camacho, O., Galan, C., Swist-Rosowska, K., Ching, R., Gamalinda, M., Karabiber, F., De La Rosa-Velazquez, I., Engist, B., Koschorz, B., Shukeir, N. et al. (2017) Major satellite repeat RNA stabilize heterochroma)n reten)on of Suv39h enzymes by RNA-nucleosome associa)on and RNA:DNA hybrid forma)on. Elife, 6. 92. Müller-Ok, K., Erdel, F., Matveeva, A., Mallm, J.-P. , R a d e m a c h e r, A . , H a h n , M . , B a u e r, C . , Z h a n g , Q., Kaltofen, S., Schoka, G. et al. (2014) Specificity, propaga)on, and memory of pericentric heterochroma)n. Mol Syst Biol, 10, 746. 93. Kudithipudi, S., Schuhmacher, M.K., Kebede, A.F. and Jeltsch, A. (2017) The SUV39H1 Protein Lysine Methyltransferase Methylates Chroma)n Proteins Involved in Heterochroma)n Forma)on and VDJ Recombina)on. ACS Chem Biol, 12, 958-968. 94. Ma, R., Zhang, Y ., Zhang, J., Zhang, P ., Liu, Z., Fan, Y ., Wang, H.T., Zhang, Z. and Zhu, B. (2024) Targe)ng pericentric non-consecu)ve mo)fs for heterochroma)n ini)a)on. Nature, 631, 678- 685. 95. Burton, A., Brochard, V., Galan, C., Ruiz-Morales, E.R., Rovira, Q., Rodriguez-Terrones, D., Kruse, K., Le Gras, S., Udayakumar, V.S., Chin, H.G. et al. (2020) Heterochroma)n establishment during early mammalian development is regulated by pericentromeric RNA and characterized by non- repressive H3K9me3. Nat Cell Biol, 22, 767-778. 96. Koo, S.J., Fernández-Montalván, A.E., Badock, V., Ok, C.J., Holton, S.J., von Ahsen, O., Toedling, J., Vikori, S., Bradner, J.E. and Gorjánácz, M. (2016) ATAD2 is an epigene)c reader of newly synthesized histone marks during DNA replica)on. Oncotarget, 7, 70323-70335. 97. Barman, S., Roy, A., Bardhan, I., Kandasamy, T., Shivani, S. and Sudhamalla, B. (2021) Insights into the Molecular Mechanisms of Histone Code Recogni)on by the BRPF3 Bromodomain. Chemistry – An Asian Journal, 16, 3404-3412. 98. Lalonde, M.-E., Avvakumov, N., Glass, K.C., Joncas, F.-H., Saksouk, N., Holliday, M., Paquet, E., Yan, K., Tong, Q., Klein, B.J. et al. (2013) Exchange of associated factors directs a switch in HBO1 acetyltransferase histone tail specificity. Genes Dev., 27, 2009-2024. 99. Lazarchuk, P ., Hernandez-Villanueva, J., Pavlova, M.N., Federa)on, A., MacCoss, M. and Sidorova, J.M. (2020) Mutual Balance of Histone Deacetylases 1 and 2 and the Acetyl Reader ATAD2 Regulates the Level of Acetyla)on of Histone H4 on Nascent Chroma)n of Human Cells. Molecular and Cellular Biology, 40, e00421-00419. 100. Vasseur, P., Tonazzini, S., Ziane, R., Camasses, A., Rando, O.J. and Radman-Livaja, M. (2016) Dynamics of Nucleosome Posi)oning Matura)on following Genomic Replica)on. Cell Reports, 16, 2651-2665. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 30 101. Stewart-Morgan, K.R., Reverón-Gómez, N. and Groth, A. (2019) Transcrip)on Restart Establishes Chroma)n Accessibility aser DNA Replica)on. Molecular Cell, 75, 284-297.e286. 102. Chen, E.S., Zhang, K., Nicolas, E., Cam, H.P ., Zofall, M. and Grewal, S.I.S. (2008) Cell cycle control of centromeric repeat transcrip)on and heterochroma)n assembly. Nature, 451, 734-737. 103. Huang, Y ., Sun, Y ., Qi, H., Jiang, Q., Li, J., Chang, M., Li, X., Shu, L., Duan, X., Wang, Y . et al. (2025) A conserved H3K14ub-driven H3K9me3 for chroma)n compartmentaliza)on. Nature, 647, 786- 797. 104. Liu, Y ., Hrit, J.A., Chomiak, A.A., Stransky, S., Hoffman, J.R., Tiedemann, R.L., Wiseman, A.K., Kariapper, L.S., Dickson, B.M., Worden, E.J. et al. (2025) DNA hypomethyla)on promotes UHRF1- and SUV39H1/H2-dependent crosstalk between H3K18ub and H3K9me3 to reinforce heterochroma)n states. Mol Cell, 85, 394-412 e312. Figure Legends Figure 1. Hat1 regulates primed enhancers in HADs. (A) Average number of H3K4me1 peaks between Hat1 KO and WT cells in HADs (top) and non-HAD LADs (bokom). Error bars represent 1 standard devia)on. P-values were calculated by Student’s t-test with n=3 biological replicates. (B) Comparison of H3K27ac as in panel A. (C) IGV browser view of chroma)n accessibility, H3K4me1, H3K27ac, and RNA-seq at the Serpin B8 locus. Figure 2. Acetyla)on changes in LADs upon Hat1 loss. (A) IGV browser view of chroma)n accessibility and the Hat1 KO/WT log2 fold-change of histone acetyla)on across chromosome 1. (B) Profile plots depic)ng the abundance of histone acetyla)on in Hat1 WT (blue) and KO (red) cells across LADs that overlap HADs (“HADs”) and LADs which do not overlap HADs (“nhLADs”). Profiles show all LADs scaled to the same size and extend 300 kb beyond each LAD border. (C) IGV browser view of a region in chromosome 4 showing loss of basal H4K5ac across LADs and gain of H4K5ac outside LAD borders in Hat1 KO cells. Figure 3. Heterochroma)n changes upon Hat1 loss. (A) Profile plots depic)ng the abundance of the specified PTM across LADs in Hat1 WT (blue) and KO (red) cells. Profiles show all LADs scaled to the same size and extend 300 kb beyond each LAD border. (B) IGV browser view of H3K9me2, H3K9me3, and H3K27me3 CUT&Tag in 3 LADS, one of which overlaps a HAD. (C) Regression plot of H3K9me2 logFC vs. H3K9me3 logFC in LADs. Each LAD is a single data point. (D) Profile plots like those in panel A, comparing HP1β in LADs that overlap HADs and LADs which do not overlap HADs. Figure 4. H3K9 histone methyltransferases in Hat1 KO and WT cells. (A) IGV browser view of HMTs and H3K9me2/3 across chromosome 2. (B) Profiles of HMT abundance across LADs. Profiles show all LADs scaled to the same size and extend 300 kb beyond each LAD border. (C) Regression plots comparing the log-transformed abundance of H3K9me3 vs. Suv39h2, Suv39h1, and Setdb1 in LADs and of H3K9me2 vs. G9a in LADs. Each LAD is a single data point. (D) IGV browser view illustra)ng changes in H3K9me2/3 and HMTs across a single LAD in Hat1 WT (colored) and KO (black). HADs overlap both ends of the LAD. (E) Profile plots of scaled LADs comparing the abundance of each HMT in Hat1 WT (blue) and KO (red) in HAD-overlapping LADs (solid lines) and non-HAD LADs (doked lines). Figure 5. ChromHMM comparison of Hat1 KO and WT chroma)n. (A) States output by chromHMM in combined WT and KO genome. The heatmap shows the probability of the mark being present in a bin of .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 31 that state. (B) Sankey plot showing state changes for 10 kb bins across the genome between WT and Hat1 KO cells. (C) Change in total genome coverage for each state (log2 KO/WT calculated from the number of bins assigned to that state in the WT and KO genomes). (D-E) Sankey plots showing state changes for 10 kb bins in LADs and HADs. Figure 6. Hat1 colocalizes with nuclear lamina proteins. (A) Volcano plot showing average detected abundance and sta)s)cal significance of proteins iden)fied by the Hat1-APEX2 construct aser filtering with the control samples. (B) PLA between two Lamin B1 an)bodies. Example images (les) and distribu)on of foci counts per cell (right) in Hat1 KO and WT cells. P-values calculated by Kolmogorov- Smirnov test. (C) Example images (les) and distribu)on of foci counts (right) of PLA between Hat1 and Lamin B1 in Hat1 KO and WT cells. P-values calculated by Kolmogorov-Smirnov test. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 32 Figure 1 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 33 Figure 2 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 34 Figure 3 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 35 Figure 4 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 36 Figure 5 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint 37 Figure 6 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 23, 2026. ; https://doi.org/10.64898/2026.03.20.713225doi: bioRxiv preprint