Disruption of the epigenetic regulator BAP1 drives chromatin remodeling leading to the emergence of cells with breast cancer stem cell properties and aberrant glycosylation

Disruption of the epigenetic regulator BAP1 drives chromatin remodeling leading to the emergence of cells with breast cancer stem cell properties and aberrant glycosylation | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results Disruption of the epigenetic regulator BAP1 drives chromatin remodeling leading to the emergence of cells with breast cancer stem cell properties and aberrant glycosylation Mariana Gomes da Silva Araujo , Aurélie Sallé , Vincent Cahais , Claire Renard , Cyrille Cuenin , Caroline Pires Poubel , Stéphane Keita , Thorsten Mosler , Christine Carreira , Gabrielle Goldman Levy , Lya Parres , Ekaterina Bourova-Flin , Sophie Rousseaux , Saadi Khochbin , Akram Ghantous , Siniša Habazin , Maja Pučić-Baković , Erika Cosset , Gordan Lauc , Zdenko Herceg , Rita Khoueiry doi: https://doi.org/10.1101/2024.12.12.628129 Mariana Gomes da Silva Araujo 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Aurélie Sallé 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Vincent Cahais 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Claire Renard 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Cyrille Cuenin 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Caroline Pires Poubel 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Stéphane Keita 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Thorsten Mosler 2 Institute of Biochemistry II, Faculty of Medicine, Goethe University Frankfurt , Frankfurt, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Christine Carreira 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Gabrielle Goldman Levy 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Lya Parres 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Ekaterina Bourova-Flin 3 Université Grenoble Alpes, CNRS UMR 5309, INSERM U1209, Institute for Advanced Biosciences , 38700, Grenoble, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sophie Rousseaux 3 Université Grenoble Alpes, CNRS UMR 5309, INSERM U1209, Institute for Advanced Biosciences , 38700, Grenoble, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Saadi Khochbin 3 Université Grenoble Alpes, CNRS UMR 5309, INSERM U1209, Institute for Advanced Biosciences , 38700, Grenoble, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Akram Ghantous 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Siniša Habazin 4 Genos Glycoscience Research Laboratory , 10000, Zagreb, Croatia Find this author on Google Scholar Find this author on PubMed Search for this author on this site Maja Pučić-Baković 4 Genos Glycoscience Research Laboratory , 10000, Zagreb, Croatia Find this author on Google Scholar Find this author on PubMed Search for this author on this site Erika Cosset 5 Cancer Research Center of Lyon - CRCL , INSERM U1052, CNRS UMR 5286, 69008, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Gordan Lauc 4 Genos Glycoscience Research Laboratory , 10000, Zagreb, Croatia 6 Faculty of Pharmacy and Biochemistry, University of Zagreb , 10000, Zagreb, Croatia Find this author on Google Scholar Find this author on PubMed Search for this author on this site Zdenko Herceg 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: hercegz{at}iarc.who.int khoueiryr{at}iarc.who.int Rita Khoueiry 1 International Agency for Research on Cancer , World Health Organization, 69007, Lyon, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: hercegz{at}iarc.who.int khoueiryr{at}iarc.who.int Abstract Full Text Info/History Metrics Supplementary material Preview PDF ABSTRACT Background Epigenetic regulator genes play critical roles in controlling cell identity and are frequently disrupted in breast cancers, suggesting a key driver role in this disease and its associated phenotypes. However, specific epigenetic drivers (epidrivers) of mammary cell plasticity and their mechanistic contributions to this phenotype are poorly characterized. Methods To identify potential epidrivers of the emergence of mesenchymal breast cancer stem cell-like phenotypes in non-tumorigenic mammary cells, we employed a CRISPR/Cas9 loss-of-function screening strategy targeting epigenetic regulator genes. This approach was followed by an in-depth validation and characterization of epigenomic, transcriptomic, proteomic and phenotypic changes resulting from the disruption of the putative epidriver gene BAP1 . Results Our investigation revealed that loss of the histone deubiquitinase BAP1 impacts cellular processes associated with breast cancer cell plasticity such as epithelial-to-mesenchymal transition (EMT) and actin cytoskeleton organization. In addition, we unveiled that BAP1 loss resulted in an overall less permissive chromatin and downregulated gene expression, impacting programs that control cellular glycosylation and leading to decreased glycan abundance and complexity. BAP1 rescue restored the expression of several deregulated genes in a catalytic activity-dependent manner, suggesting that BAP1-mediated cell identity and glycosylation regulation are largely dependent on its histone deubiquitinase activity. Conclusions Overall, our results point to BAP1 disruption as a driver of mammary cell plasticity and reveal a novel role of BAP1 as an epigenetic regulator of cellular glycosylation. BACKGROUND International wide-ranging efforts in sequencing cancer genomes and epigenomes have revealed a high frequency of alterations in genes encoding proteins that regulate the epigenome (epigenetic regulator genes, ERGs) across different cancer types 1 , 2 . This striking discovery underscored the importance of epigenetic mechanisms in cancer causation and biology. In a recent study, we showed that the disruption of ERGs may be associated with cancer hallmarks in many cancer types, supporting the notion that ERG disruption underpins rampant epigenome alterations in cancers 3 . However, the timing and stage at which the disruption of ERGs assume a primary role in tumorigenesis are poorly understood. In a cancer setting, epigenetic deregulations largely contribute to oncogenic reprogramming, enhanced cell plasticity, and maintenance of specific cancer cell phenotypes 4 , 5 . Therefore, epigenetic aberrations may enable tumorigenesis through the emergence of cancer stem cells (CSCs) or cancer initiating cells, and underlie tumor promotion and progression by conferring specific properties to cells 6 , 7 . In breast cancer (BC), subpopulations of highly plastic cells which can transition into different CSC states have been described. These breast CSCs (BCSCs) contribute to tumor growth, invasion, and resistance to treatment 8 , and are present in different stages of BC development and progression, as well as across BC molecular subtypes 9 . BCSCs may arise from differentiated breast cells or normal mammary stem cells and progenitors, owing to an acquired cell plasticity phenotype which closely associates with epithelial-to-mesenchymal transition (EMT) 10 - 12 . In addition, BCSCs with different characteristics may be simultaneously present in the same tumor or in tumors of different BC molecular subtypes, including a mesenchymal-like BCSC subpopulation characterized by CD44+ CD24-EpCAM-marker expression 13 . The high plasticity and phenotypic heterogeneity of BCSCs present challenges to their therapeutic targeting 9 , highlighting the need to further explore BCSC states, their drivers and their specific stemness-associated phenotypes. We previously observed that genetic and expression alterations in ERGs are frequent in BC patients 3 , supporting the notion that ERG deregulation may contribute to BC development. These alterations may potentially underpin the epigenome alterations described in BC tumors, such as those leading to aberrant expression of CSC and EMT-associated genes 14 - 16 . Still, most of the available data on epigenetic alterations in BC fail to capture contributions of specific ERGs to different steps of tumor development and progression, including the emergence of cells with BCSC properties. Moreover, specific epidrivers of cancer cell plasticity and BCSCs are yet to be identified and characterized in depth. In this study, we used a powerful CRISPR/Cas9 screening tool combined with an in-silico approach focused on ERGs to identify potential epidrivers contributing to the emergence of a mesenchymal BCSC-like phenotype in non-tumorigenic mammary cells. We further employed mechanistic and multi-omics approaches for an in-depth characterization of a top candidate epidriver, the deubiquitinase BAP1. With this, we gained insight into the BAP1-mediated epigenetic regulation of different cellular processes and the acquisition of BC cell plasticity. METHODS CRISPR/Cas9 library screening To identify ERG alterations contributing to a mesenchymal BCSC-like phenotype, we employed a CRISPR/Cas9 screening approach using a custom-made lentiviral library of 1,649 guide RNAs (gRNAs) targeting 426 ERGs 3 , including writers, readers and editors of epigenetic marks such as DNA methylation, histone post-translational modifications and chromatin remodeling, as well as chromatin regulators. Screening was performed in duplicates (n=2) in two independent Cas9-expressing MCF10A clones (MCF10A-Cas9 clones 2 and 11). Details on gRNA library and generation of cell lines stably expressing Cas9 cells are provided in 3 . MCF10A-Cas9 clones were transduced with the lentiviral gRNA library at a multiplicity of infection (MOI) of 0.1 (day -6). Nontargeting gRNA LentiArray CRISPR Negative Control Lentivirus (Thermo Fisher Scientific) was used as a negative control. Two days after infection (day -4), cells were treated with puromycin (1 µg/mL) to select transduced cells until day 0, when non-transduced cells cultured in parallel were dead after selection. At least 2x10 6 cells were maintained at each passage to ensure an average representation of 1,000x for each gRNA. Library-infected cells were sorted into different populations (BD FACSAria) according to cell surface marker expression on weeks 4 and 7 after puromycin selection for MCF10A clone 2 and clone 11 respectively, using antibodies for CD44 (1:400, 130-113-341, Miltenyi Biotech), CD24 (1:100, 130-112-656, Miltenyi Biotech) and EpCAM (1:200, 130-111-000, Miltenyi Biotech). To identify enriched and depleted gRNAs in each cell population assessed, we isolated genomic DNA from 2x10 6 cells (ensuring an average of 1 000x representation for each gRNA) with the AllPrep DNA/RNA Mini Kit (Qiagen), followed by PCR amplification of the gRNAs using forward (5’-CGATACAAGGCTGTTAGAGAGATA-3’) and reverse (5’-GTTGCTATTATGTCTACTATTCTTTCCC-3’) primers and NEBNext HighFidelity 2X PCR Master Mix (Illumina). Library preparation was performed with the Nextera DNA Flex Library Prep Kit (Illumina) according to manufacturer instructions and samples were sequenced on an Illumina MiSeq platform. FASTQ data generated by MiSeq were first analyzed in Galaxy using the BLASTN tool (v2.5.0) to obtain gRNA read counts, by mapping of identified gRNAs against all gRNA sequences in the CRISPR library. We next performed paired analysis comparing the gRNAs in the CD44 high EpCAM - CD24 low sorted population with gRNAs present in the bulk of cells at the day of sorting, using clone 2 and clone 11 gRNA read counts as replicates allowing for more stringency. The differential analysis of the gRNA read counts was performed with CRISPRAnalyzeR (DKFZ, Version: 1.50) 17 , using Model-based Analysis of Genome-wide CRISPR/Cas9 Knockout (MaGeCK) 18 (p-value threshold of < 0.05). Analysis of ERG mutations in breast cancer patient data Breast cancer patient data from The Cancer Genome Atlas Breast Cancer (TCGA-BRCA) database was extracted from cBioPortal ( https://www.cbioportal.org/datasets ). According to the information on receptor expression provided by the database, a total of 1,097 samples were stratified based on the expression of ER, PR and HER2 into the following subtypes: 1) ER+HER2-, 588 cases; 2) ER+HER2+, 141 cases; 3) ER-HER2+, 41 cases; and 4) TNBC, with no expression of ER, PR or HER2, 158 cases. We excluded 169 samples with unknown or conflicting expression of ER and/or HER2. The proportion of single nucleotide alterations (SNAs) in ERGs was calculated by counting the number of individuals from each BC subtype (group) which carry at least one SNA for each gene and dividing this value by the total number of individuals in the group. We considered ERGs as frequently mutated when somatic mutations in a specific gene were identified in over 1% of BC patients, allowing the discovery of potential BC epidrivers which are less frequently altered than well-known BC driver genes such as TP53 , ERRB2 and PI3KCA . These genes, although mutated in over 10% of cases, represent only a small portion of potential BC drivers 19 - 21 . Survival analysis Transcriptomic data was obtained from TCGA-BRCA, using log-transformed RNA-seq values normalized by the FPKM method [log2(1+FPKM)]. To identify an optimal gene expression threshold for BAP1 which stratifies individuals into “low” and “high” expression groups, we used the “ectopy” method as previously described 22 , 23 , considering a range of possible thresholds from 10 th to 90 th percentile. Each threshold was tested for its ability to discriminate between two groups of tumors (of low and high expression values) corresponding to significantly different disease-free survival probabilities. All thresholds were analyzed using Cox proportional hazard model between two groups of low and high expression levels, and an optimal threshold with a significant p-value lower than 0.05 was selected. For the BAP1 gene, the optimal threshold identified corresponds to the 10 th percentile of the total distribution of expression values in tumor samples. The software implemented for survival analysis is publicly available from https://github.com/epimed/ectopy . Generation of BAP1 knockout (KO) and rescue cells MCF10A-Cas9 clone 11 cells were plated at a density of 5x10 4 cells in 12-well plate wells, and transfected the next day with a pool of four BAP1 gRNAs at a final concentration of 2 µg (0.5 µg of each BAP1 gRNA), using Xfect Transfection Reagent (Takara Bio) according to the manufacturer’s instructions. gRNA sequences used are available in Supplementary Table S1. After 48h, transfected cells were selected with puromycin (1 µg/mL) until control non-transfected cells were dead. Individual KO clones were generated by sorting the bulk of antibiotic-selected cells into single cells in 96-well plates through flow cytometry (BD FACSAria), followed by amplification of the clones. Wildtype and catalytically inactive (C91S) BAP1 plasmids were obtained from Addgene (#154020 and #154021) and packaged into retroviral particles using Phoenix cells and CalPhos Mammalian Transfection Kit (BD Biosciences), as previously described 24 . MCF10A BAP1 KO cells were plated into 25 cm 2 flasks at a density of 5x10 5 cells, and infected the next day with 6mL of the collected and filtered Phoenix supernatant containing retroviral particles with either the wildtype or the mutant BAP1 construct. 24h after infection, cells were selected with 20 µg/mL of hygromycin until all non-infected control cells were dead. Validation of KOs and rescues was carried out with western blot for BAP1 expression. ATAC-sequencing Changes in chromatin accessibility were assessed with Assay for Transposase-Accessible Chromatin with high-throughput sequencing (ATAC-seq) in MCF10A-Cas9 and two clones of MCF10A-Cas9-BAP1 KO cells cultured as mammospheres, in duplicates (n=2), using the Omni-ATAC protocol as previously reported 25 . Briefly, a 45-minute transposition step using the Tagment DNA TDE1 Enzyme (Illumina) was performed on 5x10 4 viable cells per condition, followed by amplification with the NEBNext High-Fidelity 2X PCR Master Mix (New England Biolabs), using oligos available from 26 . Library DNA cleanups were performed with the DNA Clean & Concentrator-5 Kit (Zymo Research). Libraries were sequenced as 150 bp paired-end reads on Illumina NextSeq 500, using the High Output Kit v2.5 (150 cycles, ref. 20024907). Chromatin immunoprecipitation (ChIP) and ChIP-sequencing ChIP was performed in duplicates (n=2) in MCF10A wildtype (MCF10A-WT), MCF10A-Cas9 and two clones of MCF10A-Cas9-BAP1 KO cells cultured in attachment, using the iDeal ChIP-seq Kit for Histones (Diagenode) according to the manufacturer’s protocol. One million cells per IP were sonicated for 8 cycles [30 seconds “ON”, 30 seconds “OFF”] on a Bioruptor Pico (Diagenode), and immunoprecipitated with anti-H2AK119ub1 antibody (1:100 dilution; D27C4, Cell Signaling Technology). 10 ng of the H2AK119ub1 ChIPed DNA and inputs were used for library preparation with the TruSeq ChIP Sample Preparation Kit (Illumina), and samples were sequenced as 150 bp paired-end reads as described above. RNA-sequencing RNA-sequencing (RNA-seq) was performed in duplicates (n=2) in both attachment- and mammosphere-grown MCF10A-Cas9 and two clones of MCF10A-Cas9-BAP1 KO cells, using an initial amount of 2 µg or 1 µg of total RNA for attached and mammosphere-cultured cells respectively. Libraries were prepared using the KAPA Stranded mRNA-Seq Kit (Kapa Biosystems) following the manufacturer’s protocol and sequenced as 75 bp single-end reads on Illumina NextSeq 500, with the High Output Kit v2.5 (75 cycles, ref. 20024906). Bioinformatic analyses of sequencing data ATAC-seq reads were mapped to hg38 with BWA v0.7.17 27 and trimmed using Trim Galore v0.6.10 28 . Peak signals were called with MACS2 29 , using the options: –extsize 150, –shift 75, and –keep-duplicates 1, and filtered using irreproducible discovery rate (IDR) within each condition. We considered the union of resulting peaks and called reads without considering duplicated ones. Differential analysis was performed with DiffBind v3.12.0 30 , by grouping the two BAP1 KO clones together and comparing changes with MCF10A-Cas9 controls (FDR < 0.01). Read numbers per peak were normalized using the RLE method, and differential peaks were called using DESeq2’s internal method, with a false discovery rate (FDR) cutoff of < 0.01. Differential peaks were annotated using a customized annotation pipeline that included gene coordinates from GENCODE V43 31 , information on driver genes using combined data from the COSMIC Cancer Gene Census 32 and the InTOGen database 33 , 15 chromatin states for human mammary epithelial cells from the ChromHMM project 34 and Candidate cis-Regulatory Elements (CcREs) from ENCODE for mammary tumor cells (MCF-7) 35 . Estimation of differentially bound transcription factor motifs between BAP1 KOs and MCF10A-Cas9 control was performed with the TOBIAS package 36 , using the BINDetect tool with standard parameters and motif annotations from the JASPAR2024 vertebrates database 37 . BigWig signal files were inputted individually for each BAP1 KO clone and the MCF10A-Cas9 control, along with the list of ATAC-seq differential peaks (FDR |0.1| and p-value < 0.01 were considered significant. Results were plotted using the ggplot2 CRAN package, and Pearson correlation between replicates was calculated using the ggpurb CRAN package. In addition, a binned motif enrichment analysis approach was performed for the identification of differentially bound transcription factor motifs. This was achieved with the MOtif aNAlysis with Lisa (monaLisa) Bioconductor package 38 using the JASPAR2020 vertebrates database 39 , with an FDR cutoff of 0.05. ChIP-seq data was analyzed as described for ATAC-seq, with the –broad option in macs2 and considering duplicates for peak calling and read counting. Differential analysis was performed by comparing H2AK119ub1 peaks in BAP1 KOs with those identified in MCF10A-Cas9 and MCF10A-WT cells (FDR < 0.05). Custom annotation was performed as described above. RNA-seq reads were mapped to hg38 with STAR v2.7.3a 40 and counted with HTSeq v0.12.4 41 . Read counts were normalized using median of ratios transformation with DESeq2 v1.38.3 42 . Both BAP1 KO clones were grouped together for differential expression analysis, which was also performed with DESeq2. Genes were considered as differentially expressed between comparisons if FDR < 0.01. Gene set and gene ontology enrichment analyses Gene set enrichment analysis (GSEA) was performed on normalized gene counts from RNA-seq with the GSEA v4.3.2 software 43 , using the hallmark gene set collection from the Molecular Signatures Database (MSigDB) 44 and an FDR cut-off of 0.05, after filtering out low count measurements from RNA-seq (<100 counts). Gene ontology (GO) enrichment analysis of biological processes was performed with clusterProfiler v4.2.2 45 in R (v4.1.2) on differentially expressed genes from RNA-seq (FDR < 0.01) and the genes annotated closest to differential ATAC-seq peaks (FDR < 0.01), as well as on genes altered across different omics analyses. The “simplify” function was used to generate specific GO enrichment plots ( Figures 3D and 4B , Supplementary Figure S4C), to replace redundant GO terms with a representative term (similarity cutoff of 0.5, i.e., GO terms with over 50% similarity are replaced by a representative term). Total proteomics analysis Total proteomics analysis was performed in triplicates (n=3) in mammosphere-grown MCF10A-Cas9 and MCF10A-Cas9-BAP1 KO Cl1, using a total of 2 million cells per replicate. Sample preparation. Cell pellets were collected and washed 2x with PBS, then lysed in Lysis Buffer (2% SDS, 50mM Tris pH 8.5, 10mM TCEP, 40mM CAA, supplemented with protease inhibitor cocktail and phosphatase inhibitors), sonicated at 4°C in a Bioruptor Pico2 (30/30, 10 cycles) and boiled at 95°C for 10 min. Proteins were precipitated using methanol-chloroform and resuspended in 8M urea, 50mM Tris pH 8.5. Proteins were digested with 1:50 w/w LysC (Wako Chemicals) and 1:100 w/w trypsin (Promega) overnight at 37°C after dilution to a final urea concentration of 1M using 50mM Tris pH 8.5. Digested peptides were then acidified (pH 2–3) using trifluoroacetic acid (TFA) and purified using C18 SepPak columns (Waters). Desalted peptides were dried and resuspended in TMT-labeling buffer (200mM EPPS pH 8.2, 20% acetonitrile). 10μg of peptides per condition were subjected to TMT labeling with 1:2.5 peptide TMT ratio (w/w) for 1 hour at room temperature. The labeling reaction was quenched by addition of hydroxylamine to a final concentration of 0.5% and incubation at room temperature for 15 min. Successful TMT labeling was verified by mixing equimolar ratios of peptides and subjecting the mix to single shot LC-MS/MS analysis. Peptides were fractionated using high-pH liquid-chromatography on a micro-flow HPLC (Dionex U3000 RSLC, Thermo Scientific). 45µg of pooled and purified TMT labelled peptides resuspended in Solvent A (5mM ammonium-bicarbonate, 5% ACN) were separated on a C18 column (XSelect CSH, 1mm x 150mm, 3.5µm particle size; Waters) using a multistep gradient from 3-60% Solvent B (5mM ammonium-bicarbonate, 90% ACN) over 65 minutes at a flow rate of 30µl/min. Eluting peptides were collected every 43 seconds from minute 2 for 69 minutes into a total of 96 fractions, which were cross-concatenated into 24 fractions. Pooled fractions were dried in a vacuum concentrator and resuspended in 2% ACN, 0.1% TFA for LC-MS analysis. Mass spectrometry data acquisition. Tryptic peptides were analyzed on an Orbitrap Ascend coupled to a VanquishNeo (ThermoFisher Scientific) using a 25cm long, 75µm ID fused-silica column packed in house with 1.9µm C18 particles (Reprosil pur, Dr. Maisch), and kept at 50°C using an integrated column oven (Sonation). HPLC solvents consisted of 0.1% Formic acid in water (Buffer A) and 0.1% Formic acid, 80% acetonitrile in water (Buffer B). Assuming equal amounts in each fraction, 400ng of peptides were eluted by a non-linear gradient from 8% to 32% Buffer B over 77 minutes followed by a stepwise increase to 95% Buffer B in 11 minutes which was held for another 7 minutes. A synchronous precursor selection (SPS) multi-notch MS3 method was used in order to minimize ratio compression as previously described 46 . Full scan MS spectra (350-1,400m/z) were acquired with a resolution of 120,000 at m/z 200, maximum injection time of 100ms and AGC target value of 4x10 5 . The most intense precursors with a charge state between 2 and 5 per full scan were selected for fragmentation (“Top Speed” with a cycle time of 1.5 seconds) and isolated with a quadrupole isolation window of 0.7 Th. MS2 scans were performed in the Ion trap (Turbo) using a maximum injection time of 35ms, AGC target value of 10,000 and fragmented using CID with a normalized collision energy (NCE) of 35%. SPS-MS3 scans for quantification were triggered only after successful Real-time search against the human canonical reference proteome from SwissProt with the same search parameter as stated below for data processing in Proteome Discoverer. Criteria for passing the search were Xcorr: 1, dCn: 0.1 and precursor mass accuracy: 10ppm. Maximum search time was 35ms. MS3 acquisition was performed on the 10 most intense MS2 fragment ions with an isolation window of 0.7Th (MS) and 2m/z (MS2). Ions were fragmented using HCD with an NCE of 50% and analyzed in the Orbitrap with a resolution of 45,000 at m/z 200, scan range of 100-200m/z, AGC target value of 150,000 and a maximum injection time of 91ms. Repeated sequencing of already acquired precursors was limited by setting a dynamic exclusion of 60 seconds and 7ppm and advanced peak determination was deactivated. All spectra were acquired in centroid mode. Data analysis. MS raw data were analyzed using FragPipe v21.1, with MSFragger v.4.0 47 and Philosopher v.5.1.0 48 . Acquired spectra were searched against the human reference proteome (Taxonomy ID 9606) downloaded from UniProt (07/25/2024; 20,418 entries) with a precursor mass tolerance of 20ppm and fragment mass tolerance of 20ppm. Identifications were filtered to obtain false discovery rates (FDR) below 1% for both peptide spectrum matches (minimum peptide length of 7) and proteins using a target-decoy strategy. For all searches, carbamidomethylated cysteine was set as a fixed modification and oxidation of methionine and N-terminal protein acetylation as variable modifications with allowing up to 3 modifications per peptide. Strict trypsin cleavage was set as protein digestion rule. Label-free quantification was performed using IonQuant v.1.10.27 49 . Data were further processed using FragPipe Analyst 50 . Briefly, following cell lysis, total proteins were precipitated and digested, and resulting peptides were labelled with TMT buffer (200 mM EPPS pH 8.2, 20% acetonitrile). Mass spectrometry data was acquired on an Orbitrap Ascend coupled to a VanquishNeo (ThermoFisher Scientific), and identified peptides were mapped to the human reference proteome (ID 9606). Proteins were considered as differentially expressed if FDR < 0.01. Glycan profiling N -glycan profiling of MCF10A-Cas9 and MCF10A BAP1 KO cells cultured as mammospheres was performed with 1 million cells per replicate, for a total of 4 replicates per condition (n=4). Cell lysis, protein denaturation, alkylation, and enzymatic release of N-glycans . Cell pellets of MCF10A-Cas9 and MCF10A BAP1 KO cells cultured as mammospheres, containing 1 million cells each, were resuspended in 0.2% solution of RapiGest SF surfactant (Waters) in 1x PBS, with 5 µL 60 mM DTT (Sigma-Aldrich). Samples were sonicated for 5 minutes in an ultrasonic bath, followed by protein denaturation for 15 minutes at 60°C. After cooling down, 4 µL 160 mM IAA (Sigma-Aldrich) solution was added and protein alkylation was carried for 30 minutes, protected from light. Excess IAA was quenched by adding 1 µL of DTT solution. For deglycosylation, 1 U of PNGase F (Promega) in 5 µL 1x PBS was added and samples were incubated overnight at 37°C. Fluorescent N-glycan labelling and purification . Released N -glycans were labelled with procainamide (ProA, Acros Organics) via reductive amination. First, 4.32 mg ProA dissolved in 25 µL 30% glacial acetic acid in DMSO ( v / v ) was added to each sample and incubated at 65°C. Then, 4.48 mg of 2-picoline borane complex (Sigma-Aldrich) in 25 µL of the same solvent was added and incubated for another 1.5 hour. After cooling down, 1,080 µL of cold ACN was added and ProA-labelled N-glycans were purified first via Hydrophilic Interaction Liquid Chromatography Solid-Phase Extraction (HILIC-SPE) 51 and then on Porous Graphitic Carbon (PGC) as previously described 52 . N-glycan analysis . For HILIC-Ultra-performance liquid chromatography with fluorescence detection (HILIC-UPLC-FLR) analysis of N-glycans, we employed the Waters H-class UPLC system equipped with quaternary pump, sample manager and fluorescence detector (λ ex /λ em = 310/370 nm). ProA-labelled N-glycans (40 µL) were injected and separated on a Waters UPLC glycan bridged ethylene hybrid (BEH) amide chromatographic column (150 × 2.1 mm, 1.7 µm), with 100 mM ammonium formate (pH 4.4) as mobile phase A, and acetonitrile as mobile phase B. The separation method used a linear gradient of 25 to 40 % mobile phase A at the flow rate of 0.56 mL/minute in a 98-minute analytical run. Samples were maintained at 10°C before injection, and separated at 50°C. UPLC system was operated under Empower software (v3.6.1). The system was calibrated using an external standard of hydrolyzed and ProA-labelled glucose oligomers, from which the retention times for the individual N-glycans were converted to glucose units. Data processing was performed using an automatic processing method with a traditional integration algorithm, after which each chromatogram was manually corrected to maintain the same intervals of integration for all the samples. The chromatograms were all separated in the same manner into 47 peaks, and the amount of N-glycans in each peak was expressed as % of total integrated area. Visualization and statistical analyses GraphPad Prism v9.5.1 and v10.0.0 were used for the preparation of plots and statistical analyses of RT-qPCR, cell surface protein expression, pyrosequencing (plots only), proliferation, cell cycle, and metabolic assays. R v4.1.2 was used for volcano plots. ATAC-seq and H2AK119ub1 peaks were visualized on IGV (version 2.12.3), using bigwig and bed files. Description of additional methods is available from the Supplementary Information file. RESULTS CRISPR/Cas9 loss-of-function screening identifies putative epidrivers of mesenchymal breast cancer stem cells To identify key epidrivers in the emergence of mammary cells with mesenchymal-like BCSC characteristics, we employed a lentiviral CRISPR/Cas9 loss-of-function (LOF) screening approach targeting 426 ERGs in non-tumorigenic MCF10A mammary cells stably expressing Cas9 (MCF10A-Cas9) ( Figure 1A ). A mesenchymal BCSC-like population (CD44 high CD24 low EpCAM - ) was isolated from the library-infected cells, and enriched gRNAs were identified through MiSeq sequencing. We identified six ERGs whose LOF could be linked to the emergence of a mesenchymal BCSC-like phenotype (p-value < 0.05), all associated with histone post-translational modification: ASXL2 , BAP1 , CBX4 , KAT6B , KDM6A and PRDM9 ( Figure 1B , Supplementary Table S4). Among these, the BAP1 and ASXL2 genes, forming the Polycomb repressive deubiquitinase complex (PR-DUB), and the KDM6A gene, part of the COMPASS complex (complex of proteins associated with SET1), were previously described to regulate a permissive chromatin landscape 53 , 54 . Download figure Open in new tab Figure 1. BAP1 disruption is a putative epidriver of mesenchymal breast cancer stem cell-like cells. (A) CRISPR/Cas9 screening approach used for the identification of epigenetic regulator genes (ERGs) involved in the acquisition of mesenchymal breast cancer stem cell (BCSC) markers in non-tumorigenic breast cells. Adapted from Halaburkova et al. 2020 3 . gRNA: guide RNA. (B) Representation of enriched ERG gRNAs (false discovery rate [FDR] < 0.05) identified in the mesenchymal BCSC-like population of MCF10A cells infected with the ERG gRNA library compared to the bulk of cells on the day of sorting (n=2 MCF10A-Cas9 expressing clones). (C) Venn diagram of ERGs showing single nucleotide alterations in BC patients (TCGA-BRCA) of different molecular subtypes. Top 7 mutated ERGs identified in the TNBC subtype (proportion of SNAs (pSNA) > 0.019) are highlighted. (D) Kaplan-Meier analysis of disease-free survival in BC patients (TCGA-BRCA) divided in high and low BAP1 gene expression groups. * P < 0.05, ** P < 0.01. Interestingly, ASXL2 , BAP1 and PRDM9 were among the topmost frequently mutated ERGs in BC patients ( Figure 1C , Supplementary Table S5). Of note, BAP1 and PRDM9 were exclusively altered in triple negative breast cancers (TNBCs), suggesting a potential role in this cancer subtype, characterized by an enrichment in highly plastic cells and elevated expression of BCSC markers 55 , 56 . Assessment of the clinical impact of ERG expression in BC patients revealed that diminished expression of BAP1 correlates with poorer disease-free survival ( Figure 1D ), pointing to BAP1 disruption as a possible prognostic marker in BCs. Based on these findings, we hypothesized that loss of BAP1 function could be an epidriver of the emergence of mesenchymal BCSC-like cells in BC development. BAP1 is described as a tumor suppressor, and both germline and somatic mutations in this gene have been reported in different cancer types 57 , 58 . However, little is known regarding the mechanistic contributions of BAP1 deregulation to cancer cell plasticity and tumor biology in BC. BAP1 loss in non-tumorigenic breast cells enhances expression of BCSC and EMT markers To validate the role of BAP1 as a putative epidriver of breast cancer cell plasticity and the emergence of mesenchymal BCSC-like markers in mammary cells, we generated two independent MCF10A-Cas9 clones lacking BAP1 expression (BAP1 knockouts, KOs) ( Figure 2A ). When grown as mammospheres (mammary spheroids), a three-dimensional (3D) culture condition which allows enrichment of breast (cancer) stem cells 59 , both BAP1 KO clones exhibited visible changes in spheroid morphology ( Figure 2B ). While control MCF10A-Cas9 mammospheres displayed a round and compact shape, BAP1 KO spheroids exhibited a diffused and less uniform surface suggestive of a loss of contact between neighboring cells, a characteristic of mesenchymal cells. BAP1 KO mammospheres showed slightly reduced cell adhesion in their periphery, represented by detached cells at the borders, and presence of some peripheral bodies and intracytoplasmic vacuoles. These changes may be associated with disruptions in cell-to-cell communication and/or cell migration 60 ( Figure 2C , Supplementary Figure S1), in line with previous reports of alterations in the organization of three-dimensional cellular structures upon BAP1 loss 61 , 62 . Download figure Open in new tab Figure 2. BAP1 loss promotes changes in mammosphere architecture and enhances expression of BCSC and EMT markers. (A) BAP1 protein expression levels in MCF10A-Cas9 control cells (Ctrl), a heterozygous BAP1 deletion clone (HET), and two BAP1 knockout (KO) clones (Cl1 and Cl2). GAPDH levels were assessed as an internal control. (B) Representative images at 10x magnification of single, size-normalized MCF10A-Cas9 control and BAP1 KO mammospheres cultured for 72h. Scale bars: 100 µm. (C) Representative images at 20x magnification of single, size-normalized MCF10A-Cas9 control and BAP1 KO mammospheres stained with hematoxylin and eosin show mammosphere architecture differences and cellular changes between BAP1 KOs and controls. BAP1 KOs show disorganized architecture and a variable amount of intracytoplasmic vacuolization and cellular decohesion, all noted at the periphery of the mammospheres. Scale bars: 100 µm. See also Supplementary Figure S1. (D) Top: distribution of CD44 high and low populations in MCF10A-Cas9 controls and BAP1 KO cells cultured as mammospheres. Significance analysis performed by Fisher’s exact test on absolute cell counts (n=3, **** P < 0.0001). Bottom: proportion of BCSC- and epithelial-like populations based on expression of CD44, CD24 and EpCAM in MCF10A-Cas9 controls and BAP1 KO cells cultured as mammospheres. Significance analysis performed by Dunnett’s test (n=3, * P < 0.05, ** P < 0.01, ns: not significant). (E) Expression of EMT-associated genes in MCF10A-Cas9 and BAP1 KO cells cultured as mammospheres. Significance analysis performed by Dunnett’s test (n=3, * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001, ns: not significant). Next, we confirmed that BAP1 loss leads to the acquisition of mesenchymal BCSC-like markers in mammospheres, with a significant increase in CD44 expression and a striking shift in the proportion of CD44 high EpCAM - CD24 low mesenchymal BCSC-like cells (0.47% of cells in controls, 46.2% and 19.69% in BAP1 KOs), accompanied by a decrease in the CD44 low EpCAM + CD24 high epithelial-like population (19.7% of cells in controls, 1.43% and 4.43% in BAP1 KOs) ( Figure 2D , Supplementary Figure S2A). In addition, BAP1 KOs displayed changes in expression of EMT-associated genes, including decreased transcription of epithelial markers E-cadherin and EpCAM, and increased expression of mesenchymal markers N-cadherin, Vimentin, ZEB1 and ZEB2 ( Figure 2E , Supplementary Figure S2B), suggesting a role of BAP1 loss in regulating the EMT program. In line with changes in gene expression, CpG islands in the promoter regions of the E-cadherin and ZEB1 genes showed consistent differences in DNA methylation between BAP1 KOs and MCF10A-Cas9 controls, especially in mammosphere-cultured cells (Supplementary Figure S2C and D). Together, these results indicate that BAP1 loss can promote the acquisition of BCSC-like and mesenchymal features in non-tumorigenic mammary cells. BAP1 loss leads to global reduction in chromatin accessibility associated with H2AK119ub1 accumulation BAP1 is a deubiquitinating enzyme and the catalytical subunit of the PR-DUB complex, which regulates gene expression by removing the silencing histone mark H2AK119ub1 63 . Because removal of H2AK119ub1 by BAP1 prevents aberrant chromatin compaction, antagonizes Polycomb-mediated gene silencing and promotes transcription 63 , we sought to explore the impact of BAP1 loss in the chromatin accessibility of non-tumorigenic breast cells by performing ATAC-seq in mammosphere-cultured BAP1 KOs and MCF10A-Cas9 controls. We observed a clear shift in the cells’ epigenetic state following BAP1 loss (Supplementary Figure S3A), with chromatin accessibility changes observed in 36,478 sites (FDR < 0.01), which we defined as either ATAC loss (reduced accessibility, log2FC 0) ( Figure 3A ). Over 77% of differential peaks showed ATAC loss in the KOs (n=28,345), suggesting widespread reduction in chromatin accessibility following BAP1 disruption. Interestingly, a prominent fraction of the less accessible peaks upon BAP1 disruption was found in promoter regions (15.7% of ATAC loss, 3.1% of ATAC gain) ( Figure 3B , Supplementary Table S6), active transcription start sites (TSSs) (13.1% of ATAC loss, 0.5% of ATAC gain) and regions flanking TSSs (6.1% of ATAC loss, 3.5% of ATAC gain) ( Figure 3C , Supplementary Table S7), suggesting that loss of accessibility in these peaks might have a direct impact on the transcription of associated genes. Download figure Open in new tab Figure 3. BAP1 loss leads to global reduction in chromatin accessibility and H2AK119ub accumulation in mammary cells. (A) Volcano plot of ATAC-seq peaks identified in BAP1 KOs (n=4) compared to MCF10A-Cas9 controls (n=2) cultured as mammospheres. Blue and red-colored dots represent peaks with loss and gain of ATAC-seq accessibility (log2FC 0), respectively. (B) Genomic compartment and (C) chromatin state distribution of differential ATAC-seq peaks (FDR < 0.01) identified in BAP1 KOs compared to MCF10A-Cas9 cells cultured as mammospheres. Promoters are defined as regions 500 base pairs (bp) upstream of transcription start sites (TSSs). Kb: kilobases. Pc: Polycomb. (D) Gene ontology enrichment of genes associated with differential ATAC loss peaks (FDR < 0.01, log2FC < 0) falling 2kb upstream from TSSs in BAP1 KOs compared to MCF10A-Cas9 cells cultured as mammospheres. Top 10 enriched ontologies are shown (p adjusted < 0.05). (E) Differential transcription factor binding in BAP1 KO clones compared to MCF10A-Cas9 (Ctrl) cells cultured as mammospheres. Blue: 224 underrepresented motifs (differential binding score < -0.1, p 0.1, p < 0.01); and gray: motifs with no representative alteration (-0.1 ≤ differential binding score ≤ 0.1) in ATAC-seq peaks from BAP1 KOs. (F) H2AK119ub1 levels in MCF10A-Cas9 control cells (Ctrl) and two BAP1 knockout (KO) clones (Cl1 and Cl2) cultured in attachment. H3 levels were assessed as an internal control. (G) Distribution of H2AK119ub1 ChIP-seq peaks (attached cells, FDR < 0.05) in regions of ATAC loss (mammosphere cultures, FDR < 0.01, log2FC < 0) identified in BAP1 KOs compared to MCF10A-Cas9 cells. Regions within 2kb upstream and downstream of TSSs and transcription end sites (TESs), respectively, are shown. ChIP-seq reads were normalized with RPGC (reads per genome coverage) per bin method. ChIP-seq peak distribution was analyzed for MCF10A wildtype (MCF10A-WT, n=2), MCF10A-Cas9 controls (MCF10A-Cas9, n=2), and both BAP1 KO clones (BAP1 KO, n=4). Gene ontology (GO) analysis of genes associated with differential ATAC-seq peaks found within 2kb upstream of TSSs revealed reduced accessibility in genes involved in glycosylation, phospholipid synthesis and receptor signaling-associated processes ( Figure 3D ), suggesting that BAP1 may regulate cell metabolism and cellular membrane pathways. In contrast, accessible ATAC-seq peaks showed enrichment of fewer biological processes (Supplementary Figure S3B), which might be explained by the lower number of open peaks identified in regions 2kb upstream of TSSs following BAP1 loss. Next, we investigated the enrichment of transcription factor binding motifs in differentially accessible chromatin regions within 2kb upstream of TSSs. We observed a high correlation in motif enrichment between both KO clones (R=0.8, p |0.1|, p < 0.01) in MCF10A BAP1 KOs compared to controls ( Figure 3E , Supplementary Table S8). Most of the identified motifs were underrepresented in BAP1 KOs (224 motifs, differential binding score < -0.1), suggesting that BAP1 loss affects the binding of different transcription factors. Motifs with the lowest binding score in BAP1 KOs mainly include FOX transcription factors such as FOXK2, which was previously described to recruit BAP1 to chromatin 64 . On the other hand, top overrepresented motifs in BAP1 KOs (64 motifs, differential binding score > 0.1) include the TP53 and TP73 proteins, a finding confirmed with a second motif analysis tool (Supplementary Figure S3C). Together, these results point to extensive chromatin remodeling and an overall reduction in chromatin accessibility following BAP1 loss, which may impact different cellular pathways and the binding of transcription factors. We further assessed H2AK119ub1 levels in MCF10A cells following BAP1 disruption, and found an accumulation of this silencing mark upon BAP1 loss ( Figure 3F ). ChIP-seq for H2AK119ub1 revealed an overall gain in this histone mark in regions surrounding coding genes in BAP1 KOs, especially in regions overlapping with loss in chromatin accessibility ( Figure 3G , Supplementary Figure S3D). Next, we annotated enriched H2AK119ub1 peaks (FDR < 0.05) to their closest associated gene, and found that approximately 16.4% of H2AK119ub1-enriched genes were associated with ATAC loss within 2kb upstream of their TSSs, whereas only 2.6% of these genes overlapped with ATAC gain (Supplementary Figure S3E) in BAP1 KOs. Together, these data reveal that BAP1 loss results in increased levels of H2AK119ub1, which associate with loss of accessibility in regions adjacent to gene transcription, suggesting that BAP1 may directly regulate chromatin accessibility in these regions through its histone deubiquitinase function. BAP1 loss deregulates expression of genes related to metabolic and glycosylation pathways To explore the impact of BAP1 loss in gene transcription, we employed an RNA-seq approach which revealed distinct gene expression profiles between mammosphere-cultured BAP1 KOs and MCF10A-Cas9 controls (Supplementary Figure S4A). Upon BAP1 depletion, 3,925 differentially expressed genes (DEGs) (FDR < 0.01) were identified, most of them downregulated (log2FC < 0; 2,357 genes, approximately 60% of DEGs) ( Figure 4A ), in line with the overall loss in chromatin accessibility and the accumulation of H2AK119ub1 observed in BAP1 KOs. In attached cells, BAP1 KOs also exhibited a consistent downregulation in gene expression (1,536 genes out of 2,405, or 64% of DEGs) (Supplementary Figure S4B and S4C), supporting the notion that BAP1 loss leads to an overall downregulation of transcription in non-tumorigenic mammary cells. Download figure Open in new tab Figure 4. BAP1 loss deregulates gene expression of cell cycle, metabolic and glycosylation genes. (A) Volcano plot of differential expression analysis in BAP1 KOs compared to MCF10A-Cas9 control cells cultured as mammospheres (n=2 for MCF10A-Cas9 control cells and n=4 for BAP1 KOs). Blue and red-colored dots represent differentially expressed genes (DEGs, FDR < 0.01) with reduced or increased expression (log2FC 0), respectively. (B) Gene ontology enrichment of downregulated (left) and upregulated (right) DEGs (FDR < 0.01) in BAP1 KOs compared to MCF10A-Cas9 cells cultured as mammospheres. Top 10 enriched gene ontologies are shown (p adjusted < 0.05). (C) Gene set enrichment analysis (GSEA) plots of the transcriptome of BAP1 KOs compared to MCF10A-Cas9 control cells cultured as mammospheres, using the indicated gene sets. NES: normalized enrichment score. (D) Cell cycle analysis by propidium iodide staining in unsynchronized MCF10A-Cas9 and BAP1 KO mammospheres. Each bar graph shows the percentage distribution of cells in each cell cycle phase: G0-G1, S and G2-M. Significance analysis performed by Dunnett’s test (n=3, * P < 0.05, ** P < 0.01, ns: not significant). GO analysis of DEGs revealed the upregulation of ATP metabolism and mitochondrial respiration pathways in mammosphere-cultured BAP1 KOs ( Figure 4B ), indicating a possible role in the control of cellular respiration and metabolic activity. However, an assessment of mammosphere metabolism showed no alterations in mitochondrial respiration and only a slight reduction in glycolytic capacity in BAP1 KOs (Supplementary Figure S5A and B). Top downregulated biological processes upon BAP1 loss included cilium-associated and glycosylation pathways, which were also identified in cells cultured in attachment ( Figure 4B , Supplementary Figure S4C and S6). Notably, glycosylation was among the top biological processes associated with ATAC loss in BAP1 KO mammospheres, indicating that changes in expression of glycosylation genes might result from loss in chromatin accessibility. In addition, we observed downregulation of cell cycle and mitosis-associated pathways in BAP1 KO mammospheres ( Figure 4B ). Similarly, gene set enrichment analysis (GSEA) revealed negative enrichment of the E2F targets and G2M checkpoint gene sets in BAP1 KOs, both involved in DNA replication and cell cycle progression 44 ( Figure 4C ), suggesting that BAP1 loss may impact normal cell cycle progression in mammospheres, potentially coupled with EMT. Indeed, we observed an accumulation of BAP1 KO cells in the S phase accompanied by lower levels of cells in G0-G1 ( Figure 4D ), and a tendency of reduced proliferation in MCF10A BAP1 KOs cultured in attachment compared to controls (Supplementary Figure S5C), in accordance with a previous report in this same cell line 65 . Interestingly, different genes involved in cell division regulation showed reduced ATAC accessibility within 2kb upstream of their TSSs and/or associated with significant H2AK119ub1 gain in their vicinity in BAP1 KOs (Supplementary Figure S5D), suggesting that BAP1 may be involved in the transcriptional control of cell division genes through its epigenetic roles. Collectively, our transcriptomic analysis points to BAP1 as a regulator of gene expression programs associated with cell cycle progression, proliferation, glycosylation, and metabolism in non-tumorigenic mammary cells. Deregulation in gene expression following BAP1 loss is linked to BAP1’s functions as an epigenetic regulator To assess whether the changes in gene expression following BAP1 loss are associated with its role as an epigenetic regulator, we overlapped DEGs identified in BAP1 KOs with differentially accessible ATAC-seq peaks within 2kb upstream of TSSs. This revealed that reduced chromatin accessibility generally associated with downregulated gene expression, with nearly half of the downregulated genes showing concomitant loss in chromatin accessibility in BAP1 KOs, whereas gene upregulation showed less overlap with ATAC-seq signal ( Figure 5A , Supplementary Figure S7A). Importantly, the absence of changes in chromatin accessibility at most of the upregulated genes and approximately half of the downregulated genes (neutral ATAC-seq signal) may point to indirect or epigenetic regulation-independent effects of BAP1 loss. Download figure Open in new tab Figure 5. BAP1 loss-associated epigenetic deregulation impacts gene and protein expression. (A) Proportion of DEGs (FDR < 0.01) showing differential (FDR < 0.01) ATAC-seq loss (log 2 FC 0) or neutral signal (no loss or gain) within 2kb upstream of their TSSs in mammosphere-cultured BAP1 KOs compared to MCF10A-Cas9. Down: downregulated; up: upregulated. (B) Gene ontology enrichment of downregulated genes (FDR < 0.01, log 2 FC < 0) showing ATAC loss (FDR < 0.01, log 2 FC < 0) within 2kb upstream of their TSSs in mammosphere-cultured BAP1 KOs compared to MCF10A-Cas9. Top 10 enriched ontologies are shown (p adjusted < 0.05). (C) Diamond plot of glycosylation-associated DEGs (FDR < 0.01, GO: 0070085) showing changes in chromatin accessibility within 2kb upstream of their TSSs in mammosphere-cultured BAP1 KOs compared to MCF10A-Cas9 controls. Red: ATAC gain; blue: ATAC loss. y axis: RNA-seq gene expression (log2FC). (D) and (E) Genome browser snapshot of ATAC-seq and H2AK119ub1 peaks in mammosphere-cultured BAP1 KOs and MCF10A-Cas9, at the MGAT4A and ST6GALNAC1 genes. Horizontal bars represent differential ATAC-seq (FDR < 0.01) and/or ChIP-seq (FDR < 0.05) peaks. (F) Heatmap of glycosylation genes showing consistent alteration in chromatin accessibility (ATAC-seq, 2kb upstream of TSSs, left), RNA expression (RNA-seq, middle) and protein expression (proteomics, right) in BAP1 KO mammospheres compared to MCF10A-Cas9 controls. Colors and numbers on the heatmap represent the log2FC between BAP1 KOs and controls for each omics analysis (blue: ATAC loss or downregulation), and asterisks indicate that the change is significant (* FDR < 0.01). Where more than one ATAC-seq peak was annotated to a specific gene, only the closest peak to its TSS is shown in the heatmap. GO analysis of downregulated genes overlapping with ATAC loss within 2kb upstream of TSSs showed enrichment in several cellular membrane-associated pathways, including top pathways previously identified by individual analysis of RNA-seq and ATAC-seq data, such as glycosylation ( Figure 5B ). Further assessment of chromatin accessibility and gene expression changes in genes from the glycosylation pathway (GO:0070085) revealed that differentially expressed glycosylation genes, most of them downregulated in BAP1 KOs and involved in the enzymatic addition of sugars to glycan chains (glycosyltransferases), correlate with changes in ATAC-seq signal within 2kb upstream of their TSSs ( Figure 5C ). Overall, the overlap between transcriptomic and ATAC-seq data suggests a correlation between reduced chromatin accessibility and downregulated gene expression in BAP1 KO mammospheres, especially in glycosylation-associated pathways. To unravel genes and pathways that are directly regulated by BAP1’s histone deubiquitinase role, we overlapped downregulated genes, genes associated with ATAC loss (2kb upstream of TSSs), and genes associated with H2AK119ub1 gain (closest genes to H2AK119ub1 peaks) in BAP1 KO cells compared to controls. We identified 250 downregulated genes showing concomitant ATAC loss and accumulation of the silencing H2AK119ub1 mark, involved mainly in cytoskeleton and actin filament organization (Supplementary Figure S7B). The deregulation of these pathways may underpin the differences in spheroid organization and morphology displayed between controls and BAP1 KOs, and could be linked to cytoskeleton reorganization processes observed in EMT 66 . In addition, overlap of genes associated with H2AK119ub1 gain and transcription downregulation revealed enrichment of glycosylation, cilium assembly, phospholipid, and cellular membrane pathways (Supplementary Figure S7C). These included several glycosyltransferase genes, such as C1GALT1 , GALNT3 , GCNT1 , GCNT3 , MGAT4A and ST6GALNAC1 , which showed concomitant alteration in chromatin accessibility at the same regions in BAP1 KOs ( Figure 5D and 5E , Supplementary Figure S8). The impact of BAP1 loss in glycosylation-associated pathways was further confirmed at the protein level with a total proteomics approach (Supplementary Figures S9A and S9B), which revealed downregulated expression of several glycosylation proteins in BAP1 KO mammospheres (FDR < 0.01), consistent with reduced chromatin accessibility and transcription of the associated genes ( Figure 5D ). Moreover, total proteomics corroborated the changes in cell state and EMT, with the detected downregulation of epithelial markers EpCAM and MCAM and the upregulation of the mesenchymal marker Vimentin in BAP1 KOs, all accompanied by changes in chromatin accessibility and gene expression (Supplementary Figure S9C). Increased expression of the mesenchymal marker N-cadherin was also confirmed at the protein level upon BAP1 loss, although not linked with changes in chromatin accessibility. Finally, several genes involved in actin filament and cytoskeleton organization were found as consistently less accessible and downregulated across omics, including at the protein level (Supplementary Figure S9B). Altogether, our data reveal that loss of BAP1 leads to accumulation of H2AK119ub1 and reduced chromatin accessibility in numerous genes involved in cellular state, glycosylation and cytoskeleton organization, resulting in their downregulation at both transcript and protein levels and suggesting an impact in cell phenotypes. BAP1 loss disrupts glycan profiles of mammary cells Characterization of the epigenome, transcriptome and proteome landscapes in BAP1-disrupted mammospheres revealed extensive alterations in glycosylation-associated genes, mainly involving glycosyltransferases associated with O -glycan and N -glycan biosynthesis. To examine the impact of BAP1 loss in this biological process, we adopted a glycomics approach to assess the N -glycan profiles of mammosphere-cultured MCF10A-Cas9 and BAP1 KO cells using hydrophilic interaction liquid chromatography (HILIC), which revealed clear changes in glycan composition between BAP1 KOs and controls. Upon BAP1 loss, we observed a significant decrease in N -glycan complexity, represented by higher abundance of chromatographic peaks identified at earlier retention times, while conversely, complex N -glycans were more abundant in MCF10A-Cas9 controls ( Figure 6A and Supplementary Figure S10), pointing to an impact of BAP1 loss in the maintenance of glycan complexity in mammary cells. Download figure Open in new tab Figure 6. BAP1 loss disrupts glycan profiles, and its rescued activity restores phenotypes. (A) Relative abundance of chromatographic peaks identified by N -glycan profiling of mammosphere-cultured MCF10A-Cas9 and BAP1 KO cells. Significance analysis performed by Dunnett’s test (n=4, * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001, ns: not significant). Top most abundant peaks across samples are shown. GP: glycan peak. (B) O -linked N -acetylglucosamine (O-GlcNAc) expression levels in MCF10A-Cas9 controls (Ctrl) and two BAP1 knockout (KO) clones (Cl1 and Cl2) cultured as mammospheres. GAPDH levels were assessed as an internal control. (C) Top: schematic representation of the constructs used for BAP1 re-expression, containing an HA tag and hygromycin (Hygro) resistance. Representations of the wildtype (WT) and mutant BAP1 (C91S mutation) constructs are shown. Bottom: BAP1 protein expression levels in MCF10A-Cas9 controls (Ctrl), and two BAP1 knockout (KO) clones (Cl1 and Cl2) infected with BAP1 wildtype (BAP) and BAP1 C91S mutant (MUT) rescue plasmids. GAPDH levels were assessed as an internal control. (D) Left: distribution of CD44 high and low populations in BAP1 KO Cl1 and BAP1 rescue cells (wildtype: WT; and mutant: MUT) cultured as mammospheres. Significance analysis performed by Fisher’s exact test on absolute cell counts (n=2, **** P < 0.0001). Right: proportion of BCSC-like populations based on expression of CD44, CD24 and EpCAM in BAP1 KO Cl1 and BAP1 rescue cells (wildtype: WT; and mutant: MUT) cultured as mammospheres. Significance analysis performed by Dunnett’s test (n=2, * P < 0.05, ns: not significant). (E) Expression of glycosyltransferase genes C1GALT1 , GALNT3 , GCNT1 , MGAT4A and ST6GALNAC1 in MCF10A-Cas9 (Ctrl), BAP1 KO cells and BAP1 rescues (wildtype, +; and mutant, mut) cultured as mammospheres. Significance analysis performed by Dunnett’s test compared to control (Ctrl) sample (n=3, * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001, ns: not significant). Because BAP1 may interact with OGT, a glycosyltransferase which catalyzes the addition of O -linked N -acetylglucosamine ( O -GlcNAc) to proteins 67 , we assessed O -GlcNAc levels in BAP1 KOs and controls. Even though BAP1 loss did not alter the accessibility and expression of the OGT gene, BAP1 KOs exhibited reduced levels of O -GlcNAc compared to control mammospheres ( Figure 6B ), hinting to an impact of BAP1 deregulation in O -glycosylation levels. Together, these results indicate that chromatin accessibility changes resulting from BAP1 loss and the associated reduction in the expression of glycosylation genes and proteins impacts both glycan profiles and glycan abundance at a global level in non-tumorigenic mammary cells. Rescued expression of BAP1 restores BCSC marker and glycosylation gene levels in a catalytic activity-dependent manner To confirm if changes in mammary cell phenotype and glycosyltransferase gene expression are linked to BAP1’s deubiquitinase role, we restored BAP1 expression in both KO clones using either a wildtype BAP1 construct, or BAP1 with an inactivating mutation in its ubiquitin carboxy hydrolase domain (C91S) 68 ( Figure 6C ). Restoration of wildtype BAP1 led to re-established surface expression levels of the BCSC marker CD44 in mammospheres, with a decrease in the CD44 high population from over 75% to around 50% of cells, while the absence of catalytic activity failed to rescue normal CD44 levels ( Figure 6D ). The proportion of mesenchymal BCSC-like cells was also markedly reduced upon BAP1 restoration, but remained at similar levels to BAP1 KOs in the presence of mutant BAP1 ( Figure 6D ), suggesting that enhanced expression of BCSC markers upon BAP1 loss is dependent on its deubiquitinase activity. In addition, we demonstrated that the expression of several glycosyltransferase genes is reduced upon BAP1 loss in non-tumorigenic mammospheres, and that this reduction is dependent on BAP1’s catalytic activity ( Figure 6E ). Of note, BAP1 rescue also led to increased or re-established levels of different cell cycle genes which were identified as downregulated upon BAP1 loss (Supplementary Figure S11). Still, the expression of a few glycosylation and cell cycle-associated genes, such as GCNT1 and EREG , was only partially rescued in the presence of wildtype BAP1, suggesting that expression of these genes may also be co-regulated by other mechanisms. Overall, these results show that normal catalytic activity of BAP1 is needed to maintain adequate expression levels of CD44, EpCAM and CD24 markers, as well as cell cycle genes and several glycosyltransferases. DISCUSSION In this study, we applied a comprehensive in-vitro CRISPR/Cas9 screening approach in non-tumorigenic breast cells to identify potential epidrivers of the acquisition of BCSC markers, a phenotype associated with BC cell plasticity. Our approach led to the identification of BAP1 deregulation as a putative epidriver of BCSCs. Extensive validation and multi-omics characterization in mammary spheroids confirmed our observation and further revealed that BAP1 loss impacts cell morphology and phenotype and leads to broad epigenome, transcriptome, and proteome deregulation. Additionally, we linked the deubiquitinating activity of BAP1 to the control of glycosylation-associated gene expression and revealed glycomic aberrations in mammary cells as a result of BAP1 loss. This is the first report to unveil BAP1 disruption as an epigenetic driver of mammary cell plasticity and the emergence of mammary cells with mesenchymal BCSC-like markers (CD44 high CD24 low EpCAM - ) and altered expression of EMT-associated genes. This includes the reduced expression of epithelial markers EpCAM and MCAM, linked to changes in chromatin accessibility, and the upregulation of mesenchymal markers such as ZEB1 , a transcription factor that induces cell plasticity through transcriptional repression of tumor suppressor genes and activation of EMT 69 . Moreover, our observation that BAP1 is a frequently mutated ERG in TNBCs, a BC subtype enriched in BCSCs 55 , 56 , and that its reduced expression in BC tumors associates with lower disease-free survival, suggest that BAP1 expression may be linked to less aggressive phenotypes in BCs. Indeed, in different cancer types, BAP1 disruption and loss have been suggested to contribute to aggressiveness 70 , but the effects of its disruption in BC are still largely unexplored. Here, our analyses have shown that BAP1 loss alone is sufficient for the acquisition of mammary cells expressing BCSC markers, and revealed several other putative epidrivers of this phenotype, all involved in histone modification. Remarkably, three of these putative epidrivers ( BAP1 , ASXL2 and KDM6A ) counteract gene silencing promoted by Polycomb-group complexes 54 , 63 , 71 , suggesting that disruptions in this epigenetic mechanism may be implicated in mammary cell plasticity and the BCSC phenotype. BAP1 deletion in mammary cells unveiled extensive deregulation in their epigenetic profile, mainly associated with loss in chromatin accessibility, consistent with BAP1’s role as a histone modifier and transcriptional activator, and in line with previous reports of chromatin condensation and reduced accessibility upon BAP1 loss in other cell types 72 - 74 . Furthermore, we showed that increased H2AK119 ubiquitination levels following BAP1 loss largely overlapped with less accessible chromatin surrounding TSSs and transcriptional downregulation, revealing novel chromatin targets for BAP1 in mammary cells, including cell cycle, glycosylation, and cellular organization-associated genes. Chromatin accessibility changes observed following BAP1 loss also revealed a possible impact in the binding of different transcription factors to chromatin. This included overrepresentation of TP73 and TP53 motifs in BAP1 KOs, involved in regulation of stemness, self-renewal and cellular differentiation 75 , 76 , and TWIST1, a well-known inducer of EMT and stemness in tumor initiation and progression 77 . Interestingly, top underrepresented motifs in BAP1 KOs belong to the FOX family of transcription factors, involved in various functions such as regulation of cell cycle, metabolism, and stem cell maintenance 78 . One of the FOX transcription factors identified in our analysis, FOXK2, may recruit BAP1 to its target genes for the regulation of their expression through BAP1’s histone deubiquitinase function 64 , 79 , and was shown to inhibit the proliferative and invasive capacities of BC cells, suggesting a tumor suppressive role 80 . Here, the finding of extensive deregulation of transcription factor binding potential upon BAP1 loss, with increased binding of stemness-promoting factors and decreased binding of transcription factors with tumor suppressive roles, may be underpinning the shift in cell identity and the stemness-associated phenotypes observed in BAP1 KOs. Our results showing an accumulation of BAP1 KOs in the S phase of the cell cycle are consistent with previous reports linking BAP1 disruption to delayed cell cycle progression 61 , 65 , 67 , 81 , 82 . In addition, reduced proliferation is a common characteristic of mesenchymal BCSCs 13 , further implying that BAP1 loss may lead to the emergence of cells with this phenotype. Mechanistically, BAP1 may promote cell cycle progression and proliferation by H2AK119 deubiquitination of E2F target genes, resulting in their activation and transcription 57 , 58 . Moreover, partial or full restoration of cell division-associated gene expression upon rescue of catalytically active BAP1 further implicates its histone deubiquitinase function as a key mechanism in normal cell cycle regulation in mammary cells. Another striking finding from our in-depth characterization of BAP1 disruption in mammospheres was the overall deregulation of glycosylation processes, observed at the epigenome, transcriptome, proteome, and phenotypic levels. Previous studies reported that BAP1 mutations may impair its interaction with OGT and result in reduced levels of O -GlcNAc in mesothelioma and uveal melanoma 83 , and that mutant BAP1 tumors associate with a specific glycan signature in renal cell carcinoma 84 . Furthermore, we observed that BAP1-deficient tumors (BCs and pan cancer) show lower expression levels of several glycosylation pathway genes when compared to tumors exhibiting normal or low-level gain in BAP1 copy number (Supplementary Figure S12). This further supports an association between BAP1 function and the regulation of glycosylation that warrants further exploration. To our knowledge, the regulation of cellular glycosylation and glycan levels and complexity by BAP1 through epigenetic mechanisms had not been previously described, notably in mammary cells. Here, we reveal that changes in glycosylation pathways and glycan levels observed following BAP1 loss are largely linked to epigenetic and chromatin remodeling. In addition, we show rescued expression of several glycosyltransferases upon re-established BAP1 activity, further strengthening the notion that BAP1 maintains normal glycosylation levels through the epigenetic regulation of glycosylation-associated genes. Glycans are complex carbohydrate chains that can be attached to different molecules, with roles in regulation of cell metabolism, proliferation, cell adhesion, and cell signaling 85 , 86 . In cancers and cancer stem cells, aberrant glycosylation is commonly found either in the form of truncated glycosylation in early cancer development, characterized by impaired synthesis of complex glycans, or as increased glycan branching in cancer progression 87 , contributing to aggressive phenotypes, tumor progression, metastasis, and therapeutic resistance 86 , 88 . Different cancer stem cell markers, such as CD44, are extensively modified by glycosylation, and alterations in their glycan profiles may impact their stability, ligand binding and cell signaling activation 88 . The concomitant deregulation of glycosylation and the emergence of cells with BCSC markers upon BAP1 loss suggest a possible mechanistic link between these two events. However, further studies are warranted to better characterize the nature of altered glycan structures upon BAP1 loss, its impact on the expression of BCSC markers and its effect on the functions of BCSCs. Our observations place BAP1’s epigenetic activity and H2AK119ub1 removal as important aspects in the maintenance of mammary cell identity. In different cancer types, increased activity of Polycomb complex PRC1, which catalyzes H2AK119ub1, is associated with oncogenic roles, malignant transformation and stemness 89 - 92 . Our results suggest that similar to PRC1 overactivation, BAP1 loss may also contribute to cell plasticity and stemness-associated phenotypes, possibly through the accumulation of H2AK119 ubiquitination, with an impact on cell cycle and glycosylation pathways. Thus, a proper balance between BAP1 and PRC1 activity in regulating H2A ubiquitination levels may play a major role in preserving mammary cell state and in limiting a shift towards a possible oncogenic fate. CONCLUSIONS Our results reveal novel mechanistic roles for BAP1 as an epidriver of BCSC marker acquisition, which may be further explored for the understanding of cancer stem cell biology, BC initiation, and for the assessment of BAP1 disruption in aggressive breast cancers, such as TNBCs. This study may open novel avenues for exploring the effect of epigenetic drugs in the control of H2AK119 ubiquitination levels, including PRC1 inhibitors 93 , as a potential therapeutic approach to eliminate cells with cancer stem cell characteristics in breast cancers. DECLARATIONS Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Data availability High-throughput sequencing data for the CRISPR screen, ATAC-seq, ChIP-seq and RNA-seq experiments are available from Gene Expression Omnibus (GEO) (accession numbers GSE271680, GSE271681, GSE271682, GSE271683, GSE271684). Mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE repository with the dataset identifier PXD057440. Funding This work was supported by Institut National du Cancer (INCa, France) [to Z.H. and R.K.]; Fondation ARC pour la Recherche sur le Cancer (France) [to Z.H.]; La Ligue nationale contre le cancer (LNCC, France) [to R.K. and M.G.D.S.A.]; the European Fonds for regional development (REACT-EU, IWB-EFRE-program Hessen, 20008763) for the VanquishNeo, Orbitrap Ascend LC-MS systems used in this study [to T.M.]; and the Cancer ITMO (Multi-Organisation Thematic Institute of the French Alliance for Life Sciences and Health (AVIESAN)) MIC program [to E.B-F., S.R. and S.K.]. Competing interests G.L. is the founder and CEO of Genos Ltd., a private research organization specializing in high throughput glycomic analysis, and has several patents in the field. S.H. and M.P-B. are employees of Genos Ltd. Disclaimer Where authors are identified as personnel of the International Agency for Research on Cancer/World Health Organization, the authors alone are responsible for the views expressed in this article and they do not necessarily represent the decisions, policy, or views of the International Agency for Research on Cancer/World Health Organization. Authors’ contributions M.G.D.S.A. performed experiments and bioinformatic analyses, analyzed data, and wrote the manuscript; A.S. and C.Cu. performed experiments; V.C., C.R., C.P.P. and S.K. performed bioinformatic analyses of transcriptomic and epigenomic data; T.M. performed proteomic analysis; C.Ca. generated and processed histological sections; G.G.L. analyzed histological sections; L.P. generated Cas9-expressing cell lines; E.B-F., S.R. and S.K. performed survival analysis; A.G. provided support with statistical analyses; S.H., M.P-B. and G.L. performed glycan profiling; E.C. performed metabolic assays; Z.H. designed and supervised the study, and contributed to manuscript writing; R.K. designed and supervised the study, performed experiments, analyzed data, and contributed to manuscript writing; all authors reviewed the manuscript. Acknowledgements The authors would like to thank Drs. Anne Pierre-Morel, Maria Ouzounova and Rahma Benhassoun for their scientific and technical support; the Cancer Research Centre of Lyon (CRCL) flow cytometry facility for the support with cell sorting and flow cytometry experiments; the members of the Quantitative Proteomics Unit at IBC2 (Goethe University, Frankfurt), in particular Martin Adrian-Allgood and Bhavesh Parmar for technical help and Kristina Wagner for preparing LC columns; Dr. Ivan Đikić for the scientific support with proteomics experiments; and Dr. Mirjam Zeisel for kindly providing the anti-O-GlcNAc antibody. We thank Elizabeth Page and Foster Jacobs for kindly proofreading the manuscript. LIST OF ABBREVIATIONS 2D/3D two-dimensional/three-dimensional ATAC assay for transposase-accessible chromatin bp base pairs BC breast cancer BCSC breast cancer stem cell ChIP chromatin immunoprecipitation Cl clone CRISPR clustered regularly interspaced short palindromic repeats CSC cancer stem cell Ctrl control DEG differentially expressed gene ECAR extracellular acidification rate EMT epithelial-to-mesenchymal transition ERG epigenetic regulator gene FC fold change FDR false discovery rate GO gene ontology gRNA guide RNA GSEA gene set enrichment analysis H&E hematoxylin and eosin Kb kilobases KO knockout LOF loss-of-function MET mesenchymal-to-epithelial transition MOI multiplicity of infection OCR oxygen consumption rate PCA principal component analysis PR-DUB Polycomb repressive deubiquitinase PRC1/2 Polycomb repressive complex 1 or 2 RT-qPCR quantitative reverse transcription polymerase chain reaction TNBC triple negative breast cancer TSS transcription start site WT wildtype REFERENCES ↵ Plass , C. et al. Mutations in regulators of the epigenome and their connections to global chromatin patterns in cancer . Nat Rev Genet 14 , 765 – 780 ( 2013 ). doi: 10.1038/nrg3554 OpenUrl CrossRef ↵ Timp , W. & Feinberg , A. P . Cancer as a dysregulated epigenome allowing cellular growth advantage at the expense of the host . Nat Rev Cancer 13 , 497 – 510 ( 2013 ). doi: 10.1038/nrc3486 OpenUrl CrossRef PubMed Web of Science ↵ Halaburkova , A. et al. Pan-cancer multi-omics analysis and orthogonal experimental assessment of epigenetic driver genes . Genome Res 30 , 1517 – 1532 ( 2020 ). doi: 10.1101/gr.268292.120 OpenUrl Abstract / FREE Full Text ↵ Hanahan , D . Hallmarks of Cancer: New Dimensions . Cancer Discov 12 , 31 – 46 ( 2022 ). doi: 10.1158/2159-8290.CD-21-1059 OpenUrl Abstract / FREE Full Text ↵ Jones , P. A. , Ohtani , H. , Chakravarthy , A. & De Carvalho , D. D . Epigenetic therapy in immune-oncology . Nat Rev Cancer 19 , 151 – 161 ( 2019 ). doi: 10.1038/s41568-019-0109-9 OpenUrl CrossRef PubMed ↵ Wainwright , E. N. & Scaffidi , P . Epigenetics and Cancer Stem Cells: Unleashing, Hijacking, and Restricting Cellular Plasticity . Trends Cancer 3 , 372 – 386 ( 2017 ). doi: 10.1016/j.trecan.2017.04.004 OpenUrl CrossRef PubMed ↵ Chatterjee , A. , Rodger , E. J. & Eccles , M. R . Epigenetic drivers of tumourigenesis and cancer metastasis . Semin Cancer Biol 51 , 149 – 159 ( 2018 ). doi: 10.1016/j.semcancer.2017.08.004 OpenUrl CrossRef ↵ Gupta , P. B. , Pastushenko , I. , Skibinski , A. , Blanpain , C. & Kuperwasser , C . Phenotypic Plasticity: Driver of Cancer Initiation, Progression, and Therapy Resistance . Cell Stem Cell 24 , 65 – 78 ( 2019 ). doi: 10.1016/j.stem.2018.11.011 OpenUrl CrossRef PubMed ↵ Brooks , M. D. , Burness , M. L. & Wicha , M. S . Therapeutic Implications of Cellular Heterogeneity and Plasticity in Breast Cancer . Cell Stem Cell 17 , 260 – 271 ( 2015 ). doi: 10.1016/j.stem.2015.08.014 OpenUrl CrossRef PubMed ↵ Lambert , A. W. & Weinberg , R. A . Linking EMT programmes to normal and neoplastic epithelial stem cells . Nat Rev Cancer 21 , 325 – 338 ( 2021 ). doi: 10.1038/s41568-021-00332-6 OpenUrl CrossRef PubMed Bao , L. , Cardiff , R. D. , Steinbach , P. , Messer , K. S. & Ellies , L. G . Multipotent luminal mammary cancer stem cells model tumor heterogeneity . Breast Cancer Res 17 , 137 ( 2015 ). doi: 10.1186/s13058-015-0615-y OpenUrl CrossRef PubMed ↵ Morel , A. P. et al. Generation of breast cancer stem cells through epithelial-mesenchymal transition . PLoS One 3 , e2888 ( 2008 ). OpenUrl CrossRef PubMed ↵ Liu , S. et al. Breast cancer stem cells transition between epithelial and mesenchymal states reflective of their normal counterparts . Stem Cell Reports 2 , 78 – 91 ( 2014 ). doi: 10.1016/j.stemcr.2013.11.009 OpenUrl CrossRef PubMed ↵ Basse , C. & Arock , M . The increasing roles of epigenetics in breast cancer: Implications for pathogenicity, biomarkers, prevention and treatment . Int J Cancer 137 , 2785 – 2794 ( 2015 ). doi: 10.1002/ijc.29347 OpenUrl CrossRef PubMed Garcia-Martinez , L. , Zhang , Y. , Nakata , Y. , Chan , H. L. & Morey , L . Epigenetic mechanisms in breast cancer therapy and resistance . Nat Commun 12 , 1786 ( 2021 ). doi: 10.1038/s41467-021-22024-3 OpenUrl CrossRef PubMed ↵ Serrano-Gomez , S. J. , Maziveyi , M. & Alahari , S. K . Regulation of epithelial-mesenchymal transition through epigenetic and post-translational modifications . Mol Cancer 15 , 18 ( 2016 ). doi: 10.1186/s12943-016-0502-x OpenUrl CrossRef PubMed ↵ Winter , J. et al. caRpools: an R package for exploratory data analysis and documentation of pooled CRISPR/Cas9 screens . Bioinformatics 32 , 632 – 634 ( 2016 ). doi: 10.1093/bioinformatics/btv617 OpenUrl CrossRef PubMed ↵ Li , W. et al. MAGeCK enables robust identification of essential genes from genome-scale CRISPR/Cas9 knockout screens . Genome Biol 15 , 554 ( 2014 ). doi: 10.1186/s13059-014-0554-4 OpenUrl CrossRef PubMed ↵ Garraway , L. A. & Lander , E. S . Lessons from the cancer genome . Cell 153 , 17 – 37 ( 2013 ). doi: 10.1016/j.cell.2013.03.002 OpenUrl CrossRef PubMed Web of Science Pereira , B. et al. The somatic mutation profiles of 2,433 breast cancers refines their genomic and transcriptomic landscapes . Nat Commun 7 , 11479 ( 2016 ). doi: 10.1038/ncomms11479 OpenUrl CrossRef PubMed ↵ Stephens , P. J. et al. The landscape of cancer genes and mutational processes in breast cancer . Nature 486 , 400 – 404 ( 2012 ). doi: 10.1038/nature11017 OpenUrl CrossRef PubMed Web of Science ↵ Bourova-Flin , E. et al. The combined detection of Amphiregulin , Cyclin A1 and DDX20/Gemin3 expression predicts aggressive forms of oral squamous cell carcinoma. Br J Cancer 125 , 1122-1134 ( 2021 ). doi: 10.1038/s41416-021-01491-x OpenUrl CrossRef PubMed ↵ Jacquet , E. et al. Aberrant activation of five embryonic stem cell-specific genes robustly predicts a high risk of relapse in breast cancers . BMC Genomics 24 , 463 ( 2023 ). doi: 10.1186/s12864-023-09571-3 OpenUrl CrossRef ↵ Zehbe , I. et al. Human papillomavirus 16 E6 variants differ in their dysregulation of human keratinocyte differentiation and apoptosis . Virology 383 , 69 – 77 ( 2009 ). doi: 10.1016/j.virol.2008.09.036 OpenUrl CrossRef PubMed ↵ Corces , M. R. et al. An improved ATAC-seq protocol reduces background and enables interrogation of frozen tissues . Nat Methods 14 , 959 – 962 ( 2017 ). doi: 10.1038/nmeth.4396 OpenUrl CrossRef PubMed ↵ Buenrostro , J. D. , Giresi , P. G. , Zaba , L. C. , Chang , H. Y. & Greenleaf , W. J . Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position . Nat Methods 10 , 1213 – 1218 ( 2013 ). doi: 10.1038/nmeth.2688 OpenUrl CrossRef PubMed Web of Science ↵ Li , H. & Durbin , R . Fast and accurate short read alignment with Burrows-Wheeler transform . Bioinformatics 25 , 1754 – 1760 ( 2009 ). doi: 10.1093/bioinformatics/btp324 OpenUrl CrossRef PubMed Web of Science ↵ Felix Krueger , F. J. , Phil Ewels , Ebrahim Afyounian , Michael Weinstein , Benjamin Schuster-Boeckler , Gert Hulselmans , & sclamons. FelixKrueger/TrimGalore: v0.6.10 - add default decompression path (0.6.10). FelixKrueger/TrimGalore: v0.6.10 - add default decompression path (10.16.10) ( 2023 ). ↵ Gaspar , J. M . Improved peak-calling with MACS2 . bioRxiv , 496521 ( 2018 ). doi: 10.1101/496521 OpenUrl Abstract / FREE Full Text ↵ Stark R. , B. G. DiffBind : differential binding analysis of ChIP-Seq peak data . ( 2011 ). ↵ Frankish , A. et al. GENCODE reference annotation for the human and mouse genomes . Nucleic Acids Res 47 , D766 – D773 ( 2019 ). doi: 10.1093/nar/gky955 OpenUrl CrossRef PubMed ↵ Sondka , Z. et al. The COSMIC Cancer Gene Census: describing genetic dysfunction across all human cancers . Nat Rev Cancer 18 , 696 – 705 ( 2018 ). doi: 10.1038/s41568-018-0060-1 OpenUrl CrossRef PubMed ↵ Martinez-Jimenez , F. et al. A compendium of mutational cancer driver genes . Nat Rev Cancer 20 , 555 – 572 ( 2020 ). doi: 10.1038/s41568-020-0290-x OpenUrl CrossRef PubMed ↵ Ernst , J. & Kellis , M . Chromatin-state discovery and genome annotation with ChromHMM . Nat Protoc 12 , 2478 – 2492 ( 2017 ). doi: 10.1038/nprot.2017.124 OpenUrl CrossRef PubMed ↵ Consortium , E. P. et al. Expanded encyclopaedias of DNA elements in the human and mouse genomes . Nature 583 , 699 – 710 ( 2020 ). doi: 10.1038/s41586-020-2493-4 OpenUrl CrossRef PubMed ↵ Bentsen , M. et al. ATAC-seq footprinting unravels kinetics of transcription factor binding during zygotic genome activation . Nat Commun 11 , 4267 ( 2020 ). doi: 10.1038/s41467-020-18035-1 OpenUrl CrossRef PubMed ↵ Rauluseviciute , I. et al. JASPAR 2024: 20th anniversary of the open-access database of transcription factor binding profiles . Nucleic Acids Res 52 , D174 – D182 ( 2024 ). doi: 10.1093/nar/gkad1059 OpenUrl CrossRef PubMed ↵ Machlab , D. et al. monaLisa: an R/Bioconductor package for identifying regulatory motifs . Bioinformatics 38 , 2624 – 2625 ( 2022 ). doi: 10.1093/bioinformatics/btac102 OpenUrl CrossRef ↵ Fornes , O. et al. JASPAR 2020: update of the open-access database of transcription factor binding profiles . Nucleic Acids Res 48 , D87 – D92 ( 2020 ). doi: 10.1093/nar/gkz1001 OpenUrl CrossRef PubMed ↵ Dobin , A. & Gingeras , T. R . Mapping RNA-seq Reads with STAR . Curr Protoc Bioinformatics 51 , 11 14 11 – 11 14 19 ( 2015 ). doi: 10.1002/0471250953.bi1114s51 OpenUrl CrossRef ↵ Anders , S. , Pyl , P. T. & Huber , W . HTSeq--a Python framework to work with high-throughput sequencing data . Bioinformatics 31 , 166 – 169 ( 2015 ). doi: 10.1093/bioinformatics/btu638 OpenUrl CrossRef PubMed Web of Science ↵ Love , M. I. , Huber , W. & Anders , S . Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 . Genome Biol 15 , 550 ( 2014 ). doi: 10.1186/s13059-014-0550-8 OpenUrl CrossRef PubMed ↵ Subramanian , A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles . Proc Natl Acad Sci U S A 102 , 15545 – 15550 ( 2005 ). doi: 10.1073/pnas.0506580102 OpenUrl Abstract / FREE Full Text ↵ Liberzon , A. et al. The Molecular Signatures Database (MSigDB) hallmark gene set collection . Cell Syst 1 , 417 – 425 ( 2015 ). doi: 10.1016/j.cels.2015.12.004 OpenUrl CrossRef PubMed ↵ Wu , T. et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data . Innovation (Camb ) 2 , 100141 ( 2021 ). doi: 10.1016/j.xinn.2021.100141 OpenUrl CrossRef PubMed ↵ McAlister , G. C. et al. MultiNotch MS3 enables accurate, sensitive, and multiplexed detection of differential expression across cancer cell line proteomes . Anal Chem 86 , 7150 – 7158 ( 2014 ). doi: 10.1021/ac502040v OpenUrl CrossRef PubMed ↵ Kong , A. T. , Leprevost , F. V. , Avtonomov , D. M. , Mellacheruvu , D. & Nesvizhskii , A. I . MSFragger: ultrafast and comprehensive peptide identification in mass spectrometry-based proteomics . Nat Methods 14 , 513 – 520 ( 2017 ). doi: 10.1038/nmeth.4256 OpenUrl CrossRef PubMed ↵ da Veiga Leprevost , F. , et al. Philosopher: a versatile toolkit for shotgun proteomics data analysis . Nat Methods 17 , 869 – 870 ( 2020 ). doi: 10.1038/s41592-020-0912-y OpenUrl CrossRef PubMed ↵ Yu , F. , Haynes , S. E. & Nesvizhskii , A. I . IonQuant Enables Accurate and Sensitive Label-Free Quantification With FDR-Controlled Match-Between-Runs . Mol Cell Proteomics 20 , 100077 ( 2021 ). doi: 10.1016/j.mcpro.2021.100077 OpenUrl CrossRef PubMed ↵ Hsiao , Y. et al. Analysis and Visualization of Quantitative Proteomics Data Using FragPipe-Analyst . J Proteome Res 23 , 4303 – 4315 ( 2024 ). doi: 10.1021/acs.jproteome.4c00294 OpenUrl CrossRef PubMed ↵ Akmacic , I. T. et al. High-throughput glycomics: optimization of sample preparation . Biochemistry (Mosc ) 80 , 934 – 942 ( 2015 ). doi: 10.1134/S0006297915070123 OpenUrl CrossRef PubMed ↵ Klaric , T. S. , Salopek , M. , Micek , V. , Gornik Kljaic , O. & Lauc , G . Post-natal developmental changes in the composition of the rat neocortical N-glycome . Glycobiology 31 , 636 – 648 ( 2021 ). doi: 10.1093/glycob/cwaa108 OpenUrl CrossRef PubMed ↵ Scheuermann , J. C. et al. Histone H2A deubiquitinase activity of the Polycomb repressive complex PR-DUB . Nature 465 , 243 – 247 ( 2010 ). doi: 10.1038/nature08966 OpenUrl CrossRef PubMed Web of Science ↵ Wang , L. et al. Resetting the epigenetic balance of Polycomb and COMPASS function at enhancers for cancer therapy . Nat Med 24 , 758 – 769 ( 2018 ). doi: 10.1038/s41591-018-0034-6 OpenUrl CrossRef PubMed ↵ Ma , F. et al. Enriched CD44(+)/CD24(-) population drives the aggressive phenotypes presented in triple-negative breast cancer (TNBC) . Cancer Lett 353 , 153 – 159 ( 2014 ). doi: 10.1016/j.canlet.2014.06.022 OpenUrl CrossRef PubMed ↵ Park , S. Y. , Choi , J. H. & Nam , J. S . Targeting Cancer Stem Cells in Triple-Negative Breast Cancer . Cancers (Basel ) 11 ( 2019 ). doi: 10.3390/cancers11070965 OpenUrl CrossRef PubMed ↵ Louie , B. H. & Kurzrock , R . BAP1: Not just a BRCA1-associated protein . Cancer Treat Rev 90 , 102091 ( 2020 ). doi: 10.1016/j.ctrv.2020.102091 OpenUrl CrossRef ↵ Masclef , L. et al. Roles and mechanisms of BAP1 deubiquitinase in tumor suppression . Cell Death Differ 28 , 606 – 625 ( 2021 ). doi: 10.1038/s41418-020-00709-4 OpenUrl CrossRef PubMed ↵ Piscitelli , E. et al. Culture and characterization of mammary cancer stem cells in mammospheres . Methods Mol Biol 1235 , 243 – 262 ( 2015 ). doi: 10.1007/978-1-4939-1785-3_18 OpenUrl CrossRef PubMed ↵ Shubin , A. V. , Demidyuk , I. V. , Komissarov , A. A. , Rafieva , L. M. & Kostrov , S. V . Cytoplasmic vacuolization in cell death and survival . Oncotarget 7 , 55863 – 55889 ( 2016 ). doi: 10.18632/oncotarget.10150 OpenUrl CrossRef PubMed ↵ Ruland , L. et al. Organoid models of fibrolamellar carcinoma mutations reveal hepatocyte transdifferentiation through cooperative BAP1 and PRKAR2A loss . Nat Commun 14 , 2377 ( 2023 ). doi: 10.1038/s41467-023-37951-6 OpenUrl CrossRef PubMed ↵ Artegiani , B. et al. Probing the Tumor Suppressor Function of BAP1 in CRISPR-Engineered Human Liver Organoids . Cell Stem Cell 24 , 927 – 943 e926 ( 2019 ). doi: 10.1016/j.stem.2019.04.017 OpenUrl CrossRef PubMed ↵ Campagne , A. et al. BAP1 complex promotes transcription by opposing PRC1-mediated H2A ubiquitylation . Nat Commun 10 , 348 ( 2019 ). doi: 10.1038/s41467-018-08255-x OpenUrl CrossRef PubMed ↵ Okino , Y. , Machida , Y. , Frankland-Searby , S. & Machida , Y. J . BRCA1-associated protein 1 (BAP1) deubiquitinase antagonizes the ubiquitin-mediated activation of FoxK2 target genes . J Biol Chem 290 , 1580 – 1591 ( 2015 ). doi: 10.1074/jbc.M114.609834 OpenUrl Abstract / FREE Full Text ↵ Machida , Y. J. , Machida , Y. , Vashisht , A. A. , Wohlschlegel , J. A. & Dutta , A . The deubiquitinating enzyme BAP1 regulates cell growth via interaction with HCF-1 . J Biol Chem 284 , 34179 – 34188 ( 2009 ). doi: 10.1074/jbc.M109.046755 OpenUrl Abstract / FREE Full Text ↵ Dongre , A. & Weinberg , R. A . New insights into the mechanisms of epithelial-mesenchymal transition and implications for cancer . Nat Rev Mol Cell Biol 20 , 69 – 84 ( 2019 ). doi: 10.1038/s41580-018-0080-4 OpenUrl CrossRef PubMed ↵ Yu , H. et al. The ubiquitin carboxyl hydrolase BAP1 forms a ternary complex with YY1 and HCF-1 and is a critical regulator of gene expression . Mol Cell Biol 30 , 5071 – 5085 ( 2010 ). doi: 10.1128/MCB.00396-10 OpenUrl Abstract / FREE Full Text ↵ Niersch , J. et al. A BAP1 synonymous mutation results in exon skipping, loss of function and worse patient prognosis . iScience 24 , 102173 ( 2021 ). doi: 10.1016/j.isci.2021.102173 OpenUrl CrossRef ↵ Wu , H. T. et al. Oncogenic functions of the EMT-related transcription factor ZEB1 in breast cancer . J Transl Med 18 , 51 ( 2020 ). doi: 10.1186/s12967-020-02240-z OpenUrl CrossRef PubMed ↵ Carbone , M. et al. Biological Mechanisms and Clinical Significance of BAP1 Mutations in Human Cancer . Cancer Discov 10 , 1103 – 1120 ( 2020 ). doi: 10.1158/2159-8290.CD-19-1220 OpenUrl Abstract / FREE Full Text ↵ Kolovos , P. et al. PR-DUB maintains the expression of critical genes through FOXK1/2- and ASXL1/2/3-dependent recruitment to chromatin and H2AK119ub1 deubiquitination . Genome Res 30 , 1119 – 1130 ( 2020 ). doi: 10.1101/gr.261016.120 OpenUrl Abstract / FREE Full Text ↵ Conway , E. et al. BAP1 enhances Polycomb repression by counteracting widespread H2AK119ub1 deposition and chromatin condensation . Mol Cell 81 , 3526 – 3541 e3528 ( 2021 ). doi: 10.1016/j.molcel.2021.06.020 OpenUrl CrossRef PubMed Fursova , N. A. et al. BAP1 constrains pervasive H2AK119ub1 to control the transcriptional potential of the genome . Genes Dev 35 , 749 – 770 ( 2021 ). doi: 10.1101/gad.347005.120 OpenUrl Abstract / FREE Full Text ↵ Wu , Y. et al. Epigenetic and transcriptomic characterization reveals progression markers and essential pathways in clear cell renal cell carcinoma . Nat Commun 14 , 1681 ( 2023 ). doi: 10.1038/s41467-023-37211-7 OpenUrl CrossRef PubMed ↵ Soldevilla , B. , Millan , C. S. , Bonilla , F. & Dominguez , G . The TP73 complex network: ready for clinical translation in cancer? Genes Chromosomes Cancer 52 , 989 – 1006 ( 2013 ). doi: 10.1002/gcc.22095 OpenUrl CrossRef PubMed ↵ Ghatak , D. , Das Ghosh , D. & Roychoudhury , S . Cancer Stemness: p53 at the Wheel . Front Oncol 10 , 604124 ( 2020 ). doi: 10.3389/fonc.2020.604124 OpenUrl CrossRef ↵ Yu , X. et al. Molecular mechanisms of TWIST1-regulated transcription in EMT and cancer metastasis . EMBO Rep 24 , e56902 ( 2023 ). doi: 10.15252/embr.202356902 OpenUrl CrossRef PubMed ↵ Golson , M. L. & Kaestner , K. H . Fox transcription factors: from development to disease . Development 143 , 4558 – 4570 ( 2016 ). doi: 10.1242/dev.112672 OpenUrl Abstract / FREE Full Text ↵ Ji , Z . et al. The forkhead transcription factor FOXK2 acts as a chromatin targeting factor for the BAP1-containing histone deubiquitinase complex . Nucleic Acids Res 42 , 6232 – 6242 ( 2014 ). doi: 10.1093/nar/gku274 OpenUrl CrossRef PubMed Web of Science ↵ Kang , Y. , Zhang , K. , Sun , L. & Zhang , Y . Regulation and roles of FOXK2 in cancer . Front Oncol 12 , 967625 ( 2022 ). doi: 10.3389/fonc.2022.967625 OpenUrl CrossRef ↵ Daou , S. et al. The BAP1/ASXL2 Histone H2A Deubiquitinase Complex Regulates Cell Proliferation and Is Disrupted in Cancer . J Biol Chem 290 , 28643 – 28663 ( 2015 ). doi: 10.1074/jbc.M115.661553 OpenUrl Abstract / FREE Full Text ↵ Pan , H. et al. BAP1 regulates cell cycle progression through E2F1 target genes and mediates transcriptional silencing via H2A monoubiquitination in uveal melanoma cells . Int J Biochem Cell Biol 60 , 176 – 184 ( 2015 ). doi: 10.1016/j.biocel.2015.01.001 OpenUrl CrossRef PubMed ↵ Dey , A. et al. Loss of the tumor suppressor BAP1 causes myeloid transformation . Science 337 , 1541 – 1546 ( 2012 ). doi: 10.1126/science.1221711 OpenUrl Abstract / FREE Full Text ↵ Lih , T. M. , Cho , K. C. , Schnaubelt , M. , Hu , Y. & Zhang , H . Integrated glycoproteomic characterization of clear cell renal cell carcinoma . Cell Rep 42 , 112409 ( 2023 ). doi: 10.1016/j.celrep.2023.112409 OpenUrl CrossRef PubMed ↵ Klaric , T. S. et al. Human-specific features and developmental dynamics of the brain N-glycome . Sci Adv 9 , eadg2615 ( 2023 ). doi: 10.1126/sciadv.adg2615 OpenUrl CrossRef PubMed ↵ Liang , D. et al. Glycosylation in breast cancer progression and mammary development: Molecular connections and malignant transformations . Life Sci 326 , 121781 ( 2023 ). doi: 10.1016/j.lfs.2023.121781 OpenUrl CrossRef ↵ Pinho , S. S. & Reis , C. A . Glycosylation in cancer: mechanisms and clinical implications . Nat Rev Cancer 15 , 540 – 555 ( 2015 ). doi: 10.1038/nrc3982 OpenUrl CrossRef PubMed ↵ Barkeer , S. , Chugh , S. , Batra , S. K. & Ponnusamy , M. P . Glycosylation of Cancer Stem Cells: Function in Stemness, Tumorigenesis, and Metastasis . Neoplasia 20 , 813 – 825 ( 2018 ). doi: 10.1016/j.neo.2018.06.001 OpenUrl CrossRef PubMed ↵ Richly , H. , Aloia , L. & Di Croce , L . Roles of the Polycomb group proteins in stem cells and cancer . Cell Death Dis 2 , e204 ( 2011 ). doi: 10.1038/cddis.2011.84 OpenUrl CrossRef PubMed Sauvageau , M. & Sauvageau , G . Polycomb group proteins: multi-faceted regulators of somatic stem cells and cancer . Cell Stem Cell 7 , 299 – 313 ( 2010 ). doi: 10.1016/j.stem.2010.08.002 OpenUrl CrossRef PubMed Web of Science Su , W. et al. The Polycomb Repressor Complex 1 Drives Double-Negative Prostate Cancer Metastasis by Coordinating Stemness and Immune Suppression . Cancer Cell 36 , 139 – 155 e110 ( 2019 ). doi: 10.1016/j.ccell.2019.06.009 OpenUrl CrossRef PubMed ↵ van den Boom , V ., et al. Non-canonical PRC1.1 Targets Active Genes Independent of H3K27me3 and Is Essential for Leukemogenesis . Cell Rep 14 , 332 – 346 ( 2016 ). doi: 10.1016/j.celrep.2015.12.034 OpenUrl CrossRef PubMed ↵ Shukla , S. et al. Small-molecule inhibitors targeting Polycomb repressive complex 1 RING domain . Nat Chem Biol 17 , 784 – 793 ( 2021 ). doi: 10.1038/s41589-021-00815-5 OpenUrl CrossRef PubMed Back to top Previous Next Posted December 13, 2024. Download PDF Supplementary Material Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Disruption of the epigenetic regulator BAP1 drives chromatin remodeling leading to the emergence of cells with breast cancer stem cell properties and aberrant glycosylation Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Disruption of the epigenetic regulator BAP1 drives chromatin remodeling leading to the emergence of cells with breast cancer stem cell properties and aberrant glycosylation Mariana Gomes da Silva Araujo , Aurélie Sallé , Vincent Cahais , Claire Renard , Cyrille Cuenin , Caroline Pires Poubel , Stéphane Keita , Thorsten Mosler , Christine Carreira , Gabrielle Goldman Levy , Lya Parres , Ekaterina Bourova-Flin , Sophie Rousseaux , Saadi Khochbin , Akram Ghantous , Siniša Habazin , Maja Pučić-Baković , Erika Cosset , Gordan Lauc , Zdenko Herceg , Rita Khoueiry bioRxiv 2024.12.12.628129; doi: https://doi.org/10.1101/2024.12.12.628129 Share This Article: Copy Citation Tools Disruption of the epigenetic regulator BAP1 drives chromatin remodeling leading to the emergence of cells with breast cancer stem cell properties and aberrant glycosylation Mariana Gomes da Silva Araujo , Aurélie Sallé , Vincent Cahais , Claire Renard , Cyrille Cuenin , Caroline Pires Poubel , Stéphane Keita , Thorsten Mosler , Christine Carreira , Gabrielle Goldman Levy , Lya Parres , Ekaterina Bourova-Flin , Sophie Rousseaux , Saadi Khochbin , Akram Ghantous , Siniša Habazin , Maja Pučić-Baković , Erika Cosset , Gordan Lauc , Zdenko Herceg , Rita Khoueiry bioRxiv 2024.12.12.628129; doi: https://doi.org/10.1101/2024.12.12.628129 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Cancer Biology Subject Areas All Articles Animal Behavior and Cognition (8108) Biochemistry (18939) Bioengineering (15092) Bioinformatics (44958) Biophysics (22849) Cancer Biology (19967) Cell Biology (27178) Clinical Trials (138) Developmental Biology (14077) Ecology (21277) Epidemiology (2067) Evolutionary Biology (25674) Genetics (16277) Genomics (23720) Immunology (18900) Microbiology (42964) Molecular Biology (18237) Neuroscience (94447) Paleontology (709) Pathology (3014) Pharmacology and Toxicology (5148) Physiology (8216) Plant Biology (16165) Scientific Communication and Education (2106) Synthetic Biology (4610) Systems Biology (10331) Zoology (2408) window.__CF$cv$params={r:'a48d7fd5b8692229',t:'MTc5MTcxNzg3Ng==',u:'01a12ab52bf27267852829d9217f4691',ut:'1NsQPcVFlRT_bSMI.XXSdsSaQHsqtaTPk7CfgkGx6mI-1791717878-1.2.1.1-bqL4X8QeDo5kaGN1MyO5xdxvSN3M7G7bFrxUMUUTgf0q6ziD3ipyrXRPCUHj77aqCbLnGtG9VY07dFxBwAMWcec3QQ.cOOZ0Lf2vc_N15KU',i:60};(function(){if(!document.body)return;var s=document.createElement('script');s.src='/cdn-cgi/challenge-platform/scripts/precursor/main.js';document.head.appendChild(s);})();