Bromodomain Proteins Epigenetically Regulate the Mitotically Associated lncRNA MANCR in Triple Negative Breast Cancer Cells.

Long non-coding RNA (lncRNA)-mediated control of gene expression contributes to regulation of biological processes that include proliferation and phenotype, as well as compromised expression of genes that are functionally linked to cancer initiation and tumor progression. lncRNAs have emerged as novel targets and biomarkers in breast cancer. We have shown that mitotically associated lncRNA MANCR is expressed in triple-negative breast cancer (TNBC) cells and that it serves a critical role in promoting genome stability and survival in aggressive breast cancer cells. Using an siRNA strategy, we selectively depleted BRD2, BRD3, and BRD4, singly and in combination, to establish which bromodomain proteins regulate MANCR expression in TNBC cells. Our findings were confirmed by using in situ hybridization combined with immunofluorescence analysis that revealed BRD4, either alone or with BRD2 and BRD3, can support MANCR regulation of TNBC cells. Here we provide evidence for MANCR-responsive epigenetic control of super enhancers by histone modifications that are required for gene transcription to support cell survival and expression of the epithelial tumor phenotype in triple negative breast cancer cells.

Transcriptomic Analysis Pipeline (TAP) for quality control and functional assessment of transcriptomes.

Background: RNA-sequencing (RNA-seq) has revolutionized the exploration of biological mechanisms, shedding light on the roles of non-coding RNAs, including long non-coding RNAs (lncRNAs), across various biological processes, including stress responses. Despite these advancements, there remains a gap in our understanding of the implications of different RNA-seq library protocols on comprehensive lncRNA expression analysis, particularly in non-mammalian organisms. Results: In this study, we sought to bridge this knowledge gap by investigating lncRNA expression patterns in Drosophila melanogaster under thermal stress conditions. To achieve this, we conducted a comparative analysis of two RNA-seq library protocols: polyA + RNA capture and rRNA-depletion. Our approach involved the development and application of a Transcriptome Analysis Pipeline (TAP) designed to systematically assess both the technical and functional dimensions of RNA-seq, facilitating a robust comparison of these library protocols. Our findings underscore the efficacy of the polyA + protocol in capturing the majority of expressed lncRNAs within the Drosophila melanogaster transcriptome. In contrast, rRNA-depletion exhibited limited advantages in the context of D. melanogaster studies. Notably, the polyA + protocol demonstrated superior performance in terms of usable read yield and the accurate detection of splice junctions. Conclusions: Our study introduces a versatile transcriptomic analysis pipeline, TAP, designed to uniformly process RNA-seq data from any organism with a reference genome. It also highlights the significance of selecting an appropriate RNA-seq library protocol tailored to the specific research context. Background: Advances in next generation sequencing (NGS) technologies enable the comprehensive analysis of genetic sequences of organisms in a relatively cost-effective manner [1, 2]. Among these technologies, RNA-sequencing (RNA-seq) has emerged as a preeminent method to study fundamental biological mechanisms at the level of cells, tissues, and whole organisms. RNA-seq enables the detection and quantification of various RNA populations, including messenger RNA (mRNA) and various species of non-coding RNA, such as long non-coding RNA (lncRNA), as well as an assessment of features including splice junctions in RNA.

Spatiotemporal Epigenetic Control of the Histone Gene Chromatin Landscape during the Cell Cycle.

Higher-order genomic organization supports the activation of histone genes in response to cell cycle regulatory cues that epigenetically mediates stringent control of transcription at the G1/S-phase transition. Histone locus bodies (HLBs) are dynamic, non-membranous, phase-separated nuclear domains where the regulatory machinery for histone gene expression is organized and assembled to support spatiotemporal epigenetic control of histone genes. HLBs provide molecular hubs that support synthesis and processing of DNA replication-dependent histone mRNAs. These regulatory microenvironments support long-range genomic interactions among non-contiguous histone genes within a single topologically associating domain (TAD). HLBs respond to activation of the cyclin E/CDK2/NPAT/HINFP pathway at the G1/S transition. HINFP and its coactivator NPAT form a complex within HLBs that controls histone mRNA transcription to support histone protein synthesis and packaging of newly replicated DNA. Loss of HINFP compromises H4 gene expression and chromatin formation, which may result in DNA damage and impede cell cycle progression. HLBs provide a paradigm for higher-order genomic organization of a subnuclear domain that executes an obligatory cell cycle-controlled function in response to cyclin E/CDK2 signaling. Understanding the coordinately and spatiotemporally organized regulatory programs in focally defined nuclear domains provides insight into molecular infrastructure for responsiveness to cell signaling pathways that mediate biological control of growth, differentiation phenotype, and are compromised in cancer.

Spatiotemporal higher-order chromatin landscape of human histone gene clusters at histone locus bodies during the cell cycle in breast cancer progression.

Human Histone Locus Bodies (HLBs) are nuclear subdomains comprised of clustered histone genes that are coordinately regulated throughout the cell cycle. We addressed temporal-spatial higher-order genome organization for time-dependent chromatin remodeling at HLBs that supports control of cell proliferation. Proximity distances of specific genomic contacts within histone gene clusters exhibit subtle changes during the G1 phase in MCF10 breast cancer progression model cell lines. This approach directly demonstrates that the two principal histone gene regulatory proteins, HINFP (H4 gene regulator) and NPAT, localize at chromatin loop anchor-points, denoted by CTCF binding, supporting the stringent requirement for histone biosynthesis to package newly replicated DNA as chromatin. We identified a novel enhancer region located âˆ¼ 2 MB distal to histone gene sub-clusters on chromosome 6 that consistently makes genomic contacts with HLB chromatin and is bound by NPAT. During G1 progression the first DNA loops form between one of three histone gene sub-clusters bound by HINFP and the distal enhancer region. Our findings are consistent with a model that the HINFP/NPAT complex controls the formation and dynamic remodeling of higher-order genomic organization of histone gene clusters at HLBs in early to late G1 phase to support transcription of histone mRNAs in S phase.

peaksat: an R package for ChIP-seq peak saturation analysis.

BACKGROUND: Epigenomic profiling assays such as ChIP-seq have been widely used to map the genome-wide enrichment profiles of chromatin-associated proteins and posttranslational histone modifications. Sequencing depth is a key parameter in experimental design and quality control. However, due to variable sequencing depth requirements across experimental conditions, it can be challenging to determine optimal sequencing depth, particularly for projects involving multiple targets or cell types. RESULTS: We developed the peaksat R package to provide target read depth estimates for epigenomic experiments based on the analysis of peak saturation curves. We applied peaksat to establish the distinctive read depth requirements for ChIP-seq studies of histone modifications in different cell lines. Using peaksat, we were able to estimate the target read depth required per library to obtain high-quality peak calls for downstream analysis. In addition, peaksat was applied to other sequence-enrichment methods including CUT&RUN and ATAC-seq. CONCLUSION: peaksat addresses a need for researchers to make informed decisions about whether their sequencing data has been generated to an adequate depth and subsequently sufficient meaningful peaks, and failing that, how many more reads would be required per library. peaksat is applicable to other sequence-based methods that include calling peaks in their analysis.

Ikaros Regulates microRNA Networks in Acute Lymphoblastic Leukemia.

The hematopoietic transcription factor Ikaros (IKZF1) regulates normal B cell development and functions as a tumor suppressor in precursor B cell acute lymphoblastic leukemia (B-ALL). MicroRNAs (miRNAs) are small regulatory RNAs that through post-transcriptional gene regulation play critical roles in intracellular processes including cell growth in cancer. However, the role of Ikaros in the regulation of miRNA expression in developing B cells is unknown. In this study, we examined the Ikaros-regulated miRNA targets using human IKZF1-mutated Ph+ B-ALL cell lines. Inducible expression of wild-type Ikaros (the Ik1 isoform) caused B-ALL growth arrest and exit from the cell cycle. Global miRNA expression analysis revealed a total of 31 miRNAs regulated by IK1, and ChIP-seq analysis showed that Ikaros bound to several Ik1-responsive miRNA genes. Examination of the prognostic significance of miRNA expression in B-ALL indicate that the IK1-regulated miRNAs hsa-miR-26b, hsa-miR-130b and hsa-miR-4649 are significantly associated with outcome in B-ALL. Our findings establish a potential regulatory circuit between the tumor-suppressor Ikaros and the oncogenic miRNA networks in IKZF1-mutated B-ALL. These results indicate that Ikaros regulates the expression of a subset of miRNAs, of which several may contribute to B-ALL growth.

Epigenetic and transcriptome responsiveness to ER modulation by tissue selective estrogen complexes in breast epithelial and breast cancer cells.

Selective estrogen receptor modulators (SERMs), including the SERM/SERD bazedoxifene (BZA), are used to treat postmenopausal osteoporosis and may reduce breast cancer (BCa) risk. One of the most persistent unresolved questions regarding menopausal hormone therapy is compromised control of proliferation and phenotype because of short- or long-term administration of mixed-function estrogen receptor (ER) ligands. To gain insight into epigenetic effectors of the transcriptomes of hormone and BZA-treated BCa cells, we evaluated a panel of histone modifications. The impact of short-term hormone treatment and BZA on gene expression and genome-wide epigenetic profiles was examined in ERαneg mammary epithelial cells (MCF10A) and ERα+ luminal breast cancer cells (MCF7). We tested individual components and combinations of 17ß-estradiol (E2), estrogen compounds (EC10) and BZA. RNA-seq for gene expression and ChIP-seq for active (H3K4me3, H3K4ac, H3K27ac) and repressive (H3K27me3) histone modifications were performed. Our results show that the combination of BZA with E2 or EC10 reduces estrogen-mediated patterns of histone modifications and gene expression in MCF-7ERα+ cells. In contrast, BZA has minimal effects on these parameters in MCF10A mammary epithelial cells. BZA-induced changes in histone modifications in MCF7 cells are characterized by altered H3K4ac patterns, with changes at distal enhancers of ERα-target genes and at promoters of non-ERα bound proliferation-related genes. Notably, the ERα target gene GREB1 is the most sensitive to BZA treatment. Our findings provide direct mechanistic-based evidence that BZA induces epigenetic changes in E2 and EC10 mediated control of ERα regulatory programs to target distinctive proliferation gene pathways that restrain the potential for breast cancer development.

Thyroid hormone dependent transcriptional programming by TRß requires SWI/SNF chromatin remodelers.

Transcriptional regulation in response to thyroid hormone (3,5,3'-triiodo-l-thyronine, T3) is a dynamic and cell-type specific process that maintains cellular homeostasis and identity in all tissues. However, our understanding of the mechanisms of thyroid hormone receptor (TR) actions at the molecular level are actively being refined. We used an integrated genomics approach to profile and characterize the cistrome of TRß, map changes in chromatin accessibility, and capture the transcriptomic changes in response to T3 in normal human thyroid cells. There are significant shifts in TRß genomic occupancy in response to T3, which are associated with differential chromatin accessibility, and differential recruitment of SWI/SNF chromatin remodelers. We further demonstrate selective recruitment of BAF and PBAF SWI/SNF complexes to TRß binding sites, revealing novel differential functions in regulating chromatin accessibility and gene expression. Our findings highlight three distinct modes of TRß interaction with chromatin and coordination of coregulator activity.

p38 MAP Kinase Signaling in Microglia Plays a Sex-Specific Protective Role in CNS Autoimmunity and Regulates Microglial Transcriptional States.

Multiple sclerosis (MS) is an autoimmune demyelinating disease of the central nervous system, representing the leading cause of non-traumatic neurologic disease in young adults. This disease is three times more common in women, yet more severe in men, but the mechanisms underlying these sex differences remain largely unknown. MS is initiated by autoreactive T helper cells, but CNS-resident and CNS-infiltrating myeloid cells are the key proximal effector cells regulating disease pathology. We have previously shown that genetic ablation of p38α MAP kinase broadly in the myeloid lineage is protective in the autoimmune model of MS, experimental autoimmune encephalomyelitis (EAE), but only in females, and not males. To precisely define the mechanisms responsible, we used multiple genetic approaches and bone marrow chimeras to ablate p38α in microglial cells, peripheral myeloid cells, or both. Deletion of p38α in both cell types recapitulated the previous sex difference, with reduced EAE severity in females. Unexpectedly, deletion of p38α in the periphery was protective in both sexes. In contrast, deletion of p38α in microglia exacerbated EAE in males only, revealing opposing roles of p38α in microglia vs. periphery. Bulk transcriptional profiling revealed that p38α regulated the expression of distinct gene modules in male vs. female microglia. Single-cell transcriptional analysis of WT and p38α-deficient microglia isolated from the inflamed CNS revealed a diversity of complex microglial states, connected by distinct convergent transcriptional trajectories. In males, microglial p38α deficiency resulted in enhanced transition from homeostatic to disease-associated microglial states, with the downregulation of regulatory genes such as Atf3, Rgs1, Socs3, and Btg2, and increased expression of inflammatory genes such as Cd74, Trem2, and MHC class I and II genes. In females, the effect of p38α deficiency was divergent, exhibiting a unique transcriptional profile that included an upregulation of tissue protective genes, and a small subset of inflammatory genes that were also upregulated in males. Taken together, these results reveal a p38α-dependent sex-specific molecular pathway in microglia that is protective in CNS autoimmunity in males, suggesting that autoimmunity in males and females is driven by distinct cellular and molecular pathways, thus suggesting design of future sex-specific therapeutic approaches.

ssvQC: an integrated CUT&RUN quality control workflow for histone modifications and transcription factors.

OBJECTIVE: Among the different methods to profile the genome-wide patterns of transcription factor binding and histone modifications in cells and tissues, CUT&RUN has emerged as a more efficient approach that allows for a higher signal-to-noise ratio using fewer number of cells compared to ChIP-seq. The results from CUT&RUN and other related sequence enrichment assays requires comprehensive quality control (QC) and comparative analysis of data quality across replicates. While several computational tools currently exist for read mapping and analysis, a systematic reporting of data quality is lacking. Our aims were to (1) compare methods for using frozen versus fresh cells for CUT&RUN and (2) to develop an easy-to-use pipeline for assessing data quality. RESULTS: We compared a workflow for CUT&RUN with fresh and frozen samples, and present an R package called ssvQC for quality control and comparison of data quality derived from CUT&RUN and other enrichment-based sequence data. Using ssvQC, we evaluate results from different CUT&RUN protocols for transcription factors and histone modifications from fresh and frozen tissue samples. Overall, this process facilitates evaluation of data quality across datasets and permits inspection of peak calling analysis, replicate analysis of different data types. The package ssvQC is readily available at https://github.com/FrietzeLabUVM/ssvQC .

Functional Roles of Bromodomain Proteins in Cancer.

Histone acetylation is generally associated with an open chromatin configuration that facilitates many cellular processes including gene transcription, DNA repair, and DNA replication. Aberrant levels of histone lysine acetylation are associated with the development of cancer. Bromodomains represent a family of structurally well-characterized effector domains that recognize acetylated lysines in chromatin. As part of their fundamental reader activity, bromodomain-containing proteins play versatile roles in epigenetic regulation, and additional functional modules are often present in the same protein, or through the assembly of larger enzymatic complexes. Dysregulated gene expression, chromosomal translocations, and/or mutations in bromodomain-containing proteins have been correlated with poor patient outcomes in cancer. Thus, bromodomains have emerged as a highly tractable class of epigenetic targets due to their well-defined structural domains, and the increasing ease of designing or screening for molecules that modulate the reading process. Recent developments in pharmacological agents that target specific bromodomains has helped to understand the diverse mechanisms that bromodomains play with their interaction partners in a variety of chromatin processes, and provide the promise of applying bromodomain inhibitors into the clinical field of cancer treatment. In this review, we explore the expression and protein interactome profiles of bromodomain-containing proteins and discuss them in terms of functional groups. Furthermore, we highlight our current understanding of the roles of bromodomain-containing proteins in cancer, as well as emerging strategies to specifically target bromodomains, including combination therapies using bromodomain inhibitors alongside traditional therapeutic approaches designed to re-program tumorigenesis and metastasis.

Low-Level Mouse DNA in Conditioned Medium Generates False Positive Cross-Species Contamination Results in Human Organoid Cultures.

Cell line authentication is critical for preventing the use of mixed or misidentified cell lines in research. Current efforts include short tandem repeat (STR) analysis and PCR-based assays to detect mixed species cultures. Using PCR analysis with mouse-specific primers, we identified contaminating mouse DNA in growth factor conditioned medium (CM) derived from the L-WRN cell line (L-WRN CM), as well as in human organoid cultures maintained in the L-WRN CM. DNA isolated from L-WRN CM matched the L-WRN cell signature by STR analysis. Organoid lines that were positive for murine DNA by PCR were further analyzed via bulk RNA-sequencing and transcripts were aligned to the human and mouse genomes. RNA analysis failed to detect mouse-specific gene expression above background levels, suggesting no viable murine cells were present in the organoid cultures. We interpret our data to show conclusive evidence that mouse cell-derived CM can be a source of contaminating murine DNA detected in human organoid cultures, even though live, transcriptionally-active murine cells are not present. Together, our findings suggest that multiple methods may be required to authenticate human organoid or cell lines and urges cautious interpretation of DNA-based PCR cell line authentication results.

Inhibition of the RUNX1-CBFß transcription factor complex compromises mammary epithelial cell identity: a phenotype potentially stabilized by mitotic gene bookmarking.

RUNX1 has recently been shown to play an important role in determination of mammary epithelial cell identity. However, mechanisms by which loss of the RUNX1 transcription factor in mammary epithelial cells leads to epithelial-to-mesenchymal transition (EMT) are not known. Here, we report that interaction between RUNX1 and its heterodimeric partner CBFß is essential for sustaining mammary epithelial cell identity. Disruption of RUNX1-CBFß interaction, DNA binding, and association with mitotic chromosomes alters cell morphology, global protein synthesis, and phenotype-related gene expression. During interphase, RUNX1 is organized as punctate, predominantly nuclear, foci that are dynamically redistributed during mitosis, with a subset localized to mitotic chromosomes. Genome-wide RUNX1 occupancy profiles for asynchronous, mitotically enriched, and early G1 breast epithelial cells reveal RUNX1 associates with RNA Pol II-transcribed protein coding and long non-coding RNA genes and RNA Pol I-transcribed ribosomal genes critical for mammary epithelial proliferation, growth, and phenotype maintenance. A subset of these genes remains occupied by the protein during the mitosis to G1 transition. Together, these findings establish that the RUNX1-CBFß complex is required for maintenance of the normal mammary epithelial phenotype and its disruption leads to EMT. Importantly, our results suggest, for the first time, that RUNX1 mitotic bookmarking of a subset of epithelial-related genes may be an important epigenetic mechanism that contributes to stabilization of the mammary epithelial cell identity.

Expression of Heterologous OsDHAR Gene Improves Glutathione (GSH)-Dependent Antioxidant System and Maintenance of Cellular Redox Status in Synechococcus elongatus PCC 7942.

An excess of reactive oxygen species (ROS) can cause severe oxidative damage to cellular components in photosynthetic cells. Antioxidant systems, such as the glutathione (GSH) pools, regulate redox status in cells to guard against such damage. Dehydroascorbate reductase (DHAR, EC 1.8.5.1) catalyzes the glutathione-dependent reduction of oxidized ascorbate (dehydroascorbate) and contains a redox active site and glutathione binding-site. The DHAR gene is important in biological and abiotic stress responses involving reduction of the oxidative damage caused by ROS. In this study, transgenic Synechococcus elongatus PCC 7942 (TA) was constructed by cloning the Oryza sativa L. japonica DHAR (OsDHAR) gene controlled by an isopropyl ß-D-1-thiogalactopyranoside (IPTG)-inducible promoter (Ptrc) into the cyanobacterium to study the functional activities of OsDHAR under oxidative stress caused by hydrogen peroxide exposure. OsDHAR expression increased the growth of S. elongatus PCC 7942 under oxidative stress by reducing the levels of hydroperoxides and malondialdehyde (MDA) and mitigating the loss of chlorophyll. DHAR and glutathione S-transferase activity were higher than in the wild-type S. elongatus PCC 7942 (WT). Additionally, overexpression of OsDHAR in S. elongatus PCC 7942 greatly increased the glutathione (GSH)/glutathione disulfide (GSSG) ratio in the presence or absence of hydrogen peroxide. These results strongly suggest that DHAR attenuates deleterious oxidative effects via the glutathione (GSH)-dependent antioxidant system in cyanobacterial cells. The expression of heterologous OsDHAR in S. elongatus PCC 7942 protected cells from oxidative damage through a GSH-dependent antioxidant system via GSH-dependent reactions at the redox active site and GSH binding site residues during oxidative stress.

RUNX1 and RUNX2 transcription factors function in opposing roles to regulate breast cancer stem cells.

Breast cancer stem cells (BCSCs) are competent to initiate tumor formation and growth and refractory to conventional therapies. Consequently BCSCs are implicated in tumor recurrence. Many signaling cascades associated with BCSCs are critical for epithelial-to-mesenchymal transition (EMT). We developed a model system to mechanistically examine BCSCs in basal-like breast cancer using MCF10AT1 FACS sorted for CD24 (negative/low in BCSCs) and CD44 (positive/high in BCSCs). Ingenuity Pathway Analysis comparing RNA-seq on the CD24-/low versus CD24+/high MCF10AT1 indicates that the top activated upstream regulators include TWIST1, TGFß1, OCT4, and other factors known to be increased in BCSCs and during EMT. The top inhibited upstream regulators include ESR1, TP63, and FAS. Consistent with our results, many genes previously demonstrated to be regulated by RUNX factors are altered in BCSCs. The RUNX2 interaction network is the top significant pathway altered between CD24-/low and CD24+/high MCF10AT1. RUNX1 is higher in expression at the RNA level than RUNX2. RUNX3 is not expressed. While, human-specific quantitative polymerase chain reaction primers demonstrate that RUNX1 and CDH1 decrease in human MCF10CA1a cells that have grown tumors within the murine mammary fat pad microenvironment, RUNX2 and VIM increase. Treatment with an inhibitor of RUNX binding to CBFß for 5 days followed by a 7-day recovery period results in EMT suggesting that loss of RUNX1, rather than increase in RUNX2, is a driver of EMT in early stage breast cancer. Increased understanding of RUNX regulation on BCSCs and EMT will provide novel insight into therapeutic strategies to prevent recurrence.

Propionic Acid Promotes the Virulent Phenotype of Crohn's Disease-Associated Adherent-Invasive Escherichia coli.

Propionic acid (PA) is a bacterium-derived intestinal antimicrobial and immune modulator used widely in food production and agriculture. Passage of Crohn's disease-associated adherent-invasive Escherichia coli (AIEC) through a murine model, in which intestinal PA levels are increased to mimic the human intestine, leads to the recovery of AIEC with significantly increased virulence. Similar phenotypic changes are observed outside the murine model when AIEC is grown in culture with PA as the sole carbon source; such PA exposure also results in AIEC that persists at 20-fold higher levels in vivo. RNA sequencing identifies an upregulation of genes involved in biofilm formation, stress response, metabolism, membrane integrity, and alternative carbon source utilization. PA exposure also increases virulence in a number of E. coli isolates from Crohn's disease patients. Removal of PA is sufficient to reverse these phenotypic changes. Our data indicate that exposure to PA results in AIEC resistance and increased virulence in its presence.

Stromal cell protein kinase C-ß inhibition enhances chemosensitivity in B cell malignancies and overcomes drug resistance.

Overcoming drug resistance remains a key challenge to cure patients with acute and chronic B cell malignancies. Here, we describe a stromal cell-autonomous signaling pathway, which contributes to drug resistance of malignant B cells. We show that protein kinase C (PKC)-ß-dependent signals from bone marrow-derived stromal cells markedly decrease the efficacy of cytotoxic therapies. Conversely, small-molecule PKC-ß inhibitors antagonize prosurvival signals from stromal cells and sensitize tumor cells to targeted and nontargeted chemotherapy, resulting in enhanced cytotoxicity and prolonged survival in vivo. Mechanistically, stromal PKC-ß controls the expression of adhesion and matrix proteins, required for activation of phosphoinositide 3-kinases (PI3Ks) and the extracellular signal-regulated kinase (ERK)-mediated stabilization of B cell lymphoma-extra large (BCL-XL) in tumor cells. Central to the stroma-mediated drug resistance is the PKC-ß-dependent activation of transcription factor EB, regulating lysosome biogenesis and plasma membrane integrity. Stroma-directed therapies, enabled by direct inhibition of PKC-ß, enhance the effectiveness of many antileukemic therapies.

Domain-invariant features for mechanism of action prediction in a multi-cell-line drug screen.

MOTIVATION: High-content screening is an important tool in drug discovery and characterization. Often, high-content drug screens are performed on one single-cell line. Yet, a single-cell line cannot be thought of as a perfect disease model. Many diseases feature an important molecular heterogeneity. Consequently, a drug may be effective against one molecular subtype of a disease, but less so against another. To characterize drugs with respect to their effect not only on one cell line but on a panel of cell lines is therefore a promising strategy to streamline the drug discovery process. RESULTS: The contribution of this article is 2-fold. First, we investigate whether we can predict drug mechanism of action (MOA) at the molecular level without optimization of the MOA classes to the screen specificities. To this end, we benchmark a set of algorithms within a conventional pipeline, and evaluate their MOA prediction performance according to a statistically rigorous framework. Second, we extend this conventional pipeline to the simultaneous analysis of multiple cell lines, each manifesting potentially different morphological baselines. For this, we propose multi-task autoencoders, including a domain-adaptive model used to construct domain-invariant feature representations across cell lines. We apply these methods to a pilot screen of two triple negative breast cancer cell lines as models for two different molecular subtypes of the disease. AVAILABILITY AND IMPLEMENTATION: https://github.com/jcboyd/multi-cell-line or https://zenodo.org/record/2677923. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Disruption of Broad Epigenetic Domains in PDAC Cells by HAT Inhibitors.

The spreading of epigenetic domains has emerged as a distinguishing epigenomic phenotype for diverse cell types. In particular, clusters of H3K27ac- and H3K4me3-marked elements, referred to as super-enhancers, and broad H3K4me3 domains, respectively, have been linked to cell identity and disease states. Here, we characterized the broad domains from different pancreatic ductal adenocarcinoma (PDAC) cell lines that represent distinct histological grades. Our integrative genomic analysis found that human derived cell line models for distinct PDAC grades exhibit characteristic broad epigenetic features associated with gene expression patterns that are predictive of patient prognosis and provide insight into pancreatic cancer cell identity. In particular, we find that genes marked by overlapping Low-Grade broad domains correspond to an epithelial phenotype and hold potential as markers for patient stratification. We further utilize ChIP-seq to compare the effects of histone acetyltransferase (HAT) inhibitors to detect global changes in histone acetylation and methylation levels. We found that HAT inhibitors impact certain broad domains of pancreatic cancer cells. Overall, our results reveal potential roles for broad domains in cells from distinct PDAC grades and demonstrate the plasticity of particular broad epigenomic domains to epigenetic inhibitors.