xcore: an R package for inference of gene expression regulators.

BACKGROUND: Elucidating the Transcription Factors (TFs) that drive the gene expression changes in a given experiment is a common question asked by researchers. The existing methods rely on the predicted Transcription Factor Binding Site (TFBS) to model the changes in the motif activity. Such methods only work for TFs that have a motif and assume the TF binding profile is the same in all cell types. RESULTS: Given the wealth of the ChIP-seq data available for a wide range of the TFs in various cell types, we propose that gene expression modeling can be done using ChIP-seq "signatures" directly, effectively skipping the motif finding and TFBS prediction steps. We present xcore, an R package that allows TF activity modeling based on ChIP-seq signatures and the user's gene expression data. We also provide xcoredata a companion data package that provides a collection of preprocessed ChIP-seq signatures. We demonstrate that xcore leads to biologically relevant predictions using transforming growth factor beta induced epithelial-mesenchymal transition time-courses, rinderpest infection time-courses, and embryonic stem cells differentiated to cardiomyocytes time-course profiled with Cap Analysis Gene Expression. CONCLUSIONS: xcore provides a simple analytical framework for gene expression modeling using linear models that can be easily incorporated into differential expression analysis pipelines. Taking advantage of public ChIP-seq databases, xcore can identify meaningful molecular signatures and relevant ChIP-seq experiments.

Defining super-enhancers by highly ranked histone H4 multi-acetylation levels identifies transcription factors associated with glioblastoma stem-like properties.

BACKGROUND: Super-enhancers (SEs), which activate genes involved in cell-type specificity, have mainly been defined as genomic regions with top-ranked enrichment(s) of histone H3 with acetylated K27 (H3K27ac) and/or transcription coactivator(s) including a bromodomain and extra-terminal domain (BET) family protein, BRD4. However, BRD4 preferentially binds to multi-acetylated histone H4, typically with acetylated K5 and K8 (H4K5acK8ac), leading us to hypothesize that SEs should be defined by high H4K5acK8ac enrichment at least as well as by that of H3K27ac. RESULTS: Here, we conducted genome-wide profiling of H4K5acK8ac and H3K27ac, BRD4 binding, and the transcriptome by using a BET inhibitor, JQ1, in three human glial cell lines. When SEs were defined as having the top ranks for H4K5acK8ac or H3K27ac signal, 43% of H4K5acK8ac-ranked SEs were distinct from H3K27ac-ranked SEs in a glioblastoma stem-like cell (GSC) line. CRISPR-Cas9-mediated deletion of the H4K5acK8ac-preferred SEs associated with MYCN and NFIC decreased the stem-like properties in GSCs. CONCLUSIONS: Collectively, our data highlights H4K5acK8ac's utility for identifying genes regulating cell-type specificity.

Functional annotation of human long noncoding RNAs via molecular phenotyping.

Long noncoding RNAs (lncRNAs) constitute the majority of transcripts in the mammalian genomes, and yet, their functions remain largely unknown. As part of the FANTOM6 project, we systematically knocked down the expression of 285 lncRNAs in human dermal fibroblasts and quantified cellular growth, morphological changes, and transcriptomic responses using Capped Analysis of Gene Expression (CAGE). Antisense oligonucleotides targeting the same lncRNAs exhibited global concordance, and the molecular phenotype, measured by CAGE, recapitulated the observed cellular phenotypes while providing additional insights on the affected genes and pathways. Here, we disseminate the largest-to-date lncRNA knockdown data set with molecular phenotyping (over 1000 CAGE deep-sequencing libraries) for further exploration and highlight functional roles for ZNF213-AS1 and lnc-KHDC3L-2.

CAGE profiling of ncRNAs in hepatocellular carcinoma reveals widespread activation of retroviral LTR promoters in virus-induced tumors.

An increasing number of noncoding RNAs (ncRNAs) have been implicated in various human diseases including cancer; however, the ncRNA transcriptome of hepatocellular carcinoma (HCC) is largely unexplored. We used CAGE to map transcription start sites across various types of human and mouse HCCs with emphasis on ncRNAs distant from protein-coding genes. Here, we report that retroviral LTR promoters, expressed in healthy tissues such as testis and placenta but not liver, are widely activated in liver tumors. Despite HCC heterogeneity, a subset of LTR-derived ncRNAs were more than 10-fold up-regulated in the vast majority of samples. HCCs with a high LTR activity mostly had a viral etiology, were less differentiated, and showed higher risk of recurrence. ChIP-seq data show that MYC and MAX are associated with ncRNA deregulation. Globally, CAGE enabled us to build a mammalian promoter map for HCC, which uncovers a new layer of complexity in HCC genomics.

Batf2/Irf1 induces inflammatory responses in classically activated macrophages, lipopolysaccharides, and mycobacterial infection.

Basic leucine zipper transcription factor Batf2 is poorly described, whereas Batf and Batf3 have been shown to play essential roles in dendritic cell, T cell, and B cell development and regulation. Batf2 was drastically induced in IFN-γ-activated classical macrophages (M1) compared with unstimulated or IL-4-activated alternative macrophages (M2). Batf2 knockdown experiments from IFN-γ-activated macrophages and subsequent expression profiling demonstrated important roles for regulation of immune responses, inducing inflammatory and host-protective genes Tnf, Ccl5, and Nos2. Mycobacterium tuberculosis (Beijing strain HN878)-infected macrophages further induced Batf2 and augmented host-protective Batf2-dependent genes, particularly in M1, whose mechanism was suggested to be mediated through both TLR2 and TLR4 by LPS and heat-killed HN878 (HKTB) stimulation experiments. Irf1 binding motif was enriched in the promoters of Batf2-regulated genes. Coimmunoprecipitation study demonstrated Batf2 association with Irf1. Furthermore, Irf1 knockdown showed downregulation of IFN-γ- or LPS/HKTB-activated host-protective genes Tnf, Ccl5, Il12b, and Nos2. Conclusively, Batf2 is an activation marker gene for M1 involved in gene regulation of IFN-γ-activated classical macrophages, as well as LPS/HKTB-induced macrophage stimulation, possibly by Batf2/Irf1 gene induction. Taken together, these results underline the role of Batf2/Irf1 in inducing inflammatory responses in M. tuberculosis infection.

Cancers of unknown primary origin (CUP) are characterized by chromosomal instability (CIN) compared to metastasis of know origin.

BACKGROUND: Cancers of unknown primary (CUPs) constitute ~5% of all cancers. The tumors have an aggressive biological and clinical behavior. The aim of the present study has been to uncover whether CUPs exhibit distinct molecular features compared to metastases of known origin. METHODS: Employing genome wide transcriptome analysis, Linear Discriminant Analysis (LDA) and Quadratic Discriminant Analysis (QDA), we defined the putative origins of a large series of CUP and how closely related a particular CUP was to corresponding metastases of known origin. LDA predictions were subsequently used to define a universal CUP core set of differentially expressed genes, that by means of gene set enrichment analysis was exploited to depict molecular pathways characterizing CUP. RESULTS: The analyses show that CUPs are distinct from metastases of known origin. CUPs exhibit inconsistent expression of conventional cancer biomarkers and QDA derived outlier scores show that CUPs are more distantly related to their primary tumor class than corresponding metastases of known origin. Gene set enrichment analysis showed that CUPs display increased expression of genes involved in DNA damage repair and mRNA signatures of chromosome instability (CIN), indicating that CUPs are chromosome unstable compared to metastases of known origin. CONCLUSIONS: CIN may account for the uncommon clinical presentation, chemoresistance and poor outcome in patients with CUP and warrant selective diagnostic strategies and treatment.

HemaExplorer: a database of mRNA expression profiles in normal and malignant haematopoiesis.

The HemaExplorer (http://servers.binf.ku.dk/hemaexplorer) is a curated database of processed mRNA Gene expression profiles (GEPs) that provides an easy display of gene expression in haematopoietic cells. HemaExplorer contains GEPs derived from mouse/human haematopoietic stem and progenitor cells as well as from more differentiated cell types. Moreover, data from distinct subtypes of human acute myeloid leukemia is included in the database allowing researchers to directly compare gene expression of leukemic cells with those of their closest normal counterpart. Normalization and batch correction lead to full integrity of the data in the database. The HemaExplorer has comprehensive visualization interface that can make it useful as a daily tool for biologists and cancer researchers to assess the expression patterns of genes encountered in research or literature. HemaExplorer is relevant for all research within the fields of leukemia, immunology, cell differentiation and the biology of the haematopoietic system.

A statistical simulation model to guide the choices of analytical methods in arrayed CRISPR screen experiments.

An arrayed CRISPR screen is a high-throughput functional genomic screening method, which typically uses 384 well plates and has different gene knockouts in different wells. Despite various computational workflows, there is currently no systematic way to find what is a good workflow for arrayed CRISPR screening data analysis. To guide this choice, we developed a statistical simulation model that mimics the data generating process of arrayed CRISPR screening experiments. Our model is flexible and can simulate effects on phenotypic readouts of various experimental factors, such as the effect size of gene editing, as well as biological and technical variations. With two examples, we showed that the simulation model can assist making principled choice of normalization and hit calling method for the arrayed CRISPR data analysis. This simulation model is implemented in an R package and can be downloaded from Github.

Molecular profiling of high-level athlete skeletal muscle after acute endurance or resistance exercise - A systems biology approach.

OBJECTIVE: Long-term high-level exercise training leads to improvements in physical performance and multi-tissue adaptation following changes in molecular pathways. While skeletal muscle baseline differences between exercise-trained and untrained individuals have been previously investigated, it remains unclear how training history influences human multi-omics responses to acute exercise. METHODS: We recruited and extensively characterized 24 individuals categorized as endurance athletes with >15 years of training history, strength athletes or control subjects. Timeseries skeletal muscle biopsies were taken from M. vastus lateralis at three time-points after endurance or resistance exercise was performed and multi-omics molecular analysis performed. RESULTS: Our analyses revealed distinct activation differences of molecular processes such as fatty- and amino acid metabolism and transcription factors such as HIF1A and the MYF-family. We show that endurance athletes have an increased abundance of carnitine-derivates while strength athletes increase specific phospholipid metabolites compared to control subjects. Additionally, for the first time, we show the metabolite sorbitol to be substantially increased with acute exercise. On transcriptional level, we show that acute resistance exercise stimulates more gene expression than acute endurance exercise. This follows a specific pattern, with endurance athletes uniquely down-regulating pathways related to mitochondria, translation and ribosomes. Finally, both forms of exercise training specialize in diverging transcriptional directions, differentiating themselves from the transcriptome of the untrained control group. CONCLUSIONS: We identify a "transcriptional specialization effect" by transcriptional narrowing and intensification, and molecular specialization effects on metabolomic level Additionally, we performed multi-omics network and cluster analysis, providing a novel resource of skeletal muscle transcriptomic and metabolomic profiling in highly trained and untrained individuals.

Inhibition of BCL2A1 by STAT5 inactivation overcomes resistance to targeted therapies of FLT3-ITD/D835 mutant AML.

Tyrosine kinase inhibitors (TKIs) are established drugs in the therapy of FLT3-ITD mutated acute myeloid leukemia (AML). However, acquired mutations, such as D835 in the tyrosine kinase domain (FLT3-ITD/D835), can induce resistance to TKIs. A cap analysis gene expression (CAGE) technology revealed that the gene expression of BCL2A1 transcription start sites was increased in primary AML cells bearing FLT3-ITD/D835 compared to FLT3-ITD. Overexpression of BCL2A1 attenuated the sensitivity to quizartinib, a type II TKI, and venetoclax, a selective BCL2 inhibitor, in AML cell lines. However, a type I TKI, gilteritinib, inhibited the expression of BCL2A1 through inactivation of STAT5 and alleviated TKI resistance of FLT3-ITD/D835. The combination of gilteritinib and venetoclax showed synergistic effects in the FLT3-ITD/D835 positive AML cells. The promoter region of BCL2A1 contains a BRD4 binding site. Thus, the blockade of BRD4 with a BET inhibitor (CPI-0610) downregulated BCL2A1 in FLT3-mutated AML cells and extended profound suppression of FLT3-ITD/D835 mutant cells. Therefore, we propose that BCL2A1 has the potential to be a novel therapeutic target in treating FLT3-ITD/D835 mutated AML.

Inducing human retinal pigment epithelium-like cells from somatic tissue.

Regenerative medicine relies on basic research outcomes that are only practical when cost effective. The human eyeball requires the retinal pigment epithelium (RPE) to interface the neural retina and the choroid at large. Millions of people suffer from age-related macular degeneration (AMD), a blinding multifactor genetic disease among RPE degradation pathologies. Recently, autologous pluripotent stem-cell-derived RPE cells were prohibitively expensive due to time; therefore, we developed a faster reprogramming system. We stably induced RPE-like cells (iRPE) from human fibroblasts (Fibs) by conditional overexpression of both broad plasticity and lineage-specific transcription factors (TFs). iRPE cells displayed critical RPE benchmarks and significant in vivo integration in transplanted retinas. Herein, we detail the iRPE system with comprehensive single-cell RNA sequencing (scRNA-seq) profiling to interpret and characterize its best cells. We anticipate that our system may enable robust retinal cell induction for basic research and affordable autologous human RPE tissue for regenerative cell therapy.

Differential expression of cellular microRNAs in HPV 11, -16, and -45 transfected cells.

Human papillomaviruses (HPVs) are highly prevalent giving rise to both benign and malignant lesions why they are classified as high- and low-risk viruses. In this study we selected one low-risk (HPV 11) and two high-risk (HPV 16 and -45) types for genomewide miRNA analysis to investigate possible common and distinct features in the expression profiles. For this purpose we developed a cell culture model system in HaCaT cells for expression of the viral genomes under standardized conditions. We identified 25 miRNAs which were differentially regulated in two or three HPV types where 13 miRNAs were in common for all three types. Among the miRNAs identified, miR-125a-5p, miR-129-3p, miR-363, and miR-145 are related to human cancers. Noteworthy, miR-145 is found upregulated in the miRNA profiles of both high-risk HPV types. For selected differentially expressed miRNAs in HPV 16 predicted miRNA target transcript involved in signal transduction, RNA splicing and tumor invasive growth were validated by qRT-PCR. In addition, our results imply that the early 3' untranslated region (3'UTR) of the three HPV genomes were not a target for miRNA regulation.

The microRNA molecular signature of atypic and common acquired melanocytic nevi: differential expression of miR-125b and let-7c.

MicroRNAs (miRNAs) are small non-coding RNAs, which regulate gene expression through base pairing with mRNA and which are crucially involved in carcinogenesis (the so-called oncomiRs). We compared the miRNA signature between acquired melanocytic nevi showing clinical atypia (atypic nevi, AN) and common acquired nevi (common nevi, CN). We obtained miRNA profiles from 41 biopsies (22 AN and 19 CN) and showed that AN could be differentiated from CN on the basis of the expression of 36 miRNAs (false discovery rate <0.05). OncomiRs were present in this group, and we further confirmed the differential expression of miR-125b and let-7c by qRT-PCR. Our data suggest that miRNAs are functionally involved in the pathogenesis of nevi and possibly malignant melanoma.

Development of p53 knockout U87MG cell line for unbiased drug delivery testing system using CRISPR-Cas9 and transcriptomic analysis.

Antibody-drug conjugates offers many advantages as a drug delivery platform that allows for highly specific targeting of cell types and genes. Ideally, testing the efficacy of these systems requires two cell types to be different only in the gene targeted by the drug, with the rest of the cellular machinery unchanged, in order to minimize other potential differences from obscuring the effects of the drug. In this study, we created multiple variants of U87MG cells with targeted mutation in the TP53 gene using the CRISPR-Cas9 system, and determined that their major transcriptional differences stem from the loss of p53 function. Using the transcriptome data, we predicted which mutant clones would have less divergent phenotypes from the wild type and thereby serve as the best candidates to be used as drug delivery testing platforms. Further in vitro and in vivo assays of cell morphology, proliferation rate and target antigen-mediated uptake supported our predictions. Based on the combined analysis results, we successfully selected the best qualifying mutant clone. This study serves as proof-of-principle of the approach and paves the way for extending to additional cell types and target genes.

Differential expression of cellular microRNAs in HPV-11 transfected cells. An analysis by three different array platforms and qRT-PCR.

Human papillomavirus type 11 (HPV-11) infects the genital and the respiratory tract leading to condylomas and respiratory papillomatosis. HPV infections are restricted to epithelial tissue and the progression through the virus lifecycle is tightly coordinated to the differentiation of the host cell. The changes of cellular microRNAs by HPV-11 gene expression were investigated in a cell culture model of HaCaT cells transfected with HPV-11, with the goal of understanding which cellular processes were affected by the virus. Human microRNA profiling was conducted on three different array platform systems and because very few microRNAs (miR-663, -638, -149* and -92b*) were consistently found in all three array data sets we performed extensive statistical analyses of the array data and the qRT-PCR validation. We assume that the most reliable differentially expressed microRNAs are the ones identified by more than one array platform. We also show that TaqMan® qRT-PCR validation is of limited use for less abundant microRNAs.

Structural profiles of human miRNA families from pairwise clustering.

UNLABELLED: MicroRNAs (miRNAs) are a group of small, approximately 21 nt long, riboregulators inhibiting gene expression at a post-transcriptional level. Their most distinctive structural feature is the foldback hairpin of their precursor pre-miRNAs. Even though each pre-miRNA deposited in miRBase has its secondary structure already predicted, little is known about the patterns of structural conservation among pre-miRNAs. We address this issue by clustering the human pre-miRNA sequences based on pairwise, sequence and secondary structure alignment using FOLDALIGN, followed by global multiple alignment of obtained clusters by WAR. As a result, the common secondary structure was successfully determined for four FOLDALIGN clusters: the RF00027 structural family of the Rfam database and three clusters with previously undescribed consensus structures. AVAILABILITY: http://genome.ku.dk/resources/mirclust

Reproducible pattern of microRNA in normal human skin.

MicroRNAs (miRNAs) regulate cell growth, differentiation and apoptosis via specific targeting of messenger RNA (mRNA). Aberrant mRNA expression contributes to pathological processes such as carcinogenesis. To take advantage of miRNA profiling in skin disease it is essential to investigate miRNA expression pattern in normal human skin. Here we investigated miRNA expression profiles from skin biopsies of 8 healthy volunteers taken from sun protected and mildly photo damaged skin using the modified protocol for miRNA extraction. We were able to show a constant pattern of miRNA expression between different individuals. We did not find any significant differences in miRNA expression between sun protected and mildly photodamaged skin. These results may be valuable for future design of studies on miRNA expression in skin disease.

A simple procedure for routine RNA extraction and miRNA array analyses from a single thyroid in vivo fine needle aspirate.

CONTEXT: microRNA (miRNA) expression profiling and classification of tissue obtained from fine-needle aspirates (FNA) could be a major improvement of the preoperative diagnosis of thyroid nodules. OBJECTIVE: Before this can be clinically implemented, a robust and non-toxic method for obtaining sufficient quantity and quality of RNA from single in vivo FNA has to be established. RNAlater is a non-toxic RNA-stabilizing agent. However, due to the high density of RNAlater, pelleting of the tissue samples is difficult, and results in low recovery of RNA that is insufficient for subsequent miRNA array expression analysis. We therefore developed a simple centrifugation method for capturing tissue stored in RNAlater on a 0.45-µm filter. DESIGN: FNA from 24 patients with a solitary cold thyroid nodule was stored in Trizol, liquid nitrogen, or RNAlater. The tissue stored in RNAlater was either pelleted by centrifugation or captured on the 0.45-µm filters. RNA was extracted using the Trizol method. To validate results, FNA from additional 30 patients were analyzed based on the modified RNAlater protocol. MAIN OUTCOME: Capturing FNA tissue samples on the filters increased the RNA yield 10 fold and maintained RNA purity, permitting miRNA array expression profiling and allowing comparable levels of known miRNA-clusters regardless of preservation technique. Results were confirmed in an additional 30 patients. CONCLUSION: The modified RNAlater protocol is well suited for isolating RNA from single thyroid in vivo FNA in a clinical setting. Furthermore this permits shipping of FNA samples at room temperature from peripheral centers to a centralized array core facility.

JQ1 affects BRD2-dependent and independent transcription regulation without disrupting H4-hyperacetylated chromatin states.

The bromodomain and extra-terminal domain (BET) proteins are promising drug targets for cancer and immune diseases. However, BET inhibition effects have been studied more in the context of bromodomain-containing protein 4 (BRD4) than BRD2, and the BET protein association to histone H4-hyperacetylated chromatin is not understood at the genome-wide level. Here, we report transcription start site (TSS)-resolution integrative analyses of ChIP-seq and transcriptome profiles in human non-small cell lung cancer (NSCLC) cell line H23. We show that di-acetylation at K5 and K8 of histone H4 (H4K5acK8ac) co-localizes with H3K27ac and BRD2 in the majority of active enhancers and promoters, where BRD2 has a stronger association with H4K5acK8ac than H3K27ac. Although BET inhibition by JQ1 led to complete reduction of BRD2 binding to chromatin, only local changes of H4K5acK8ac levels were observed, suggesting that recruitment of BRD2 does not influence global histone H4 hyperacetylation levels. This finding supports a model in which recruitment of BET proteins via histone H4 hyperacetylation is predominant over hyperacetylation of histone H4 by BET protein-associated acetyltransferases. In addition, we found that a remarkable number of BRD2-bound genes, including MYC and its downstream target genes, were transcriptionally upregulated upon JQ1 treatment. Using BRD2-enriched sites and transcriptional activity analysis, we identified candidate transcription factors potentially involved in the JQ1 response in BRD2-dependent and -independent manner.