Overview of the Knowledge Management Center for Illuminating the Druggable Genome

The Knowledge Management Center (KMC) for the Illuminating the Druggable Genome (IDG) project aims to aggregate, update, and articulate protein-centric data knowledge for the entire human proteome, with emphasis on the understudied proteins from the three IDG protein families. KMC collates and analyzes data from over 70 resources to compile the Target Central Resource Database (TCRD), which is the web-based informatics platform (Pharos). These data include experimental, computational, and text-mined information on protein structures, compound interactions, and disease and phenotype associations. Based on this knowledge, proteins are classified into different Target Development Levels (TDLs) for identification of understudied targets. Additional work by the KMC focuses on enriching target knowledge and producing DrugCentral and other data visualization tools for expanding investigation of understudied targets.

Precision information extraction for rare disease epidemiology at scale

BACKGROUND

The United Nations recently made a call to address the challenges of an estimated 300 million persons worldwide living with a rare disease through the collection, analysis, and dissemination of disaggregated data. Epidemiologic Information (EI) regarding prevalence and incidence data of rare diseases is sparse and current paradigms of identifying, extracting, and curating EI rely upon time-intensive, error-prone manual processes. With these limitations, a clear understanding of the variation in epidemiology and outcomes for rare disease patients is hampered. This challenges the public health of rare diseases patients through a lack of information necessary to prioritize research, policy decisions, therapeutic development, and health system allocations.

METHODS

In this study, we developed a newly curated epidemiology corpus for Named Entity Recognition (NER), a deep learning framework, and a novel rare disease epidemiologic information pipeline named EpiPipeline4RD consisting of a web interface and Restful API. For the corpus creation, we programmatically gathered a representative sample of rare disease epidemiologic abstracts, utilized weakly-supervised machine learning techniques to label the dataset, and manually validated the labeled dataset. For the deep learning framework development, we fine-tuned our dataset and adapted the BioBERT model for NER. We measured the performance of our BioBERT model for epidemiology entity recognition quantitatively with precision, recall, and F1 and qualitatively through a comparison with Orphanet. We demonstrated the ability for our pipeline to gather, identify, and extract epidemiology information from rare disease abstracts through three case studies.

RESULTS

We developed a deep learning model to extract EI with overall F1 scores of 0.817 and 0.878, evaluated at the entity-level and token-level respectively, and which achieved comparable qualitative results to Orphanet's collection paradigm. Additionally, case studies of the rare diseases Classic homocystinuria, GRACILE syndrome, Phenylketonuria demonstrated the adequate recall of abstracts with epidemiology information, high precision of epidemiology information extraction through our deep learning model, and the increased efficiency of EpiPipeline4RD compared to a manual curation paradigm.

CONCLUSIONS

EpiPipeline4RD demonstrated high performance of EI extraction from rare disease literature to augment manual curation processes. This automated information curation paradigm will not only effectively empower development of the NIH Genetic and Rare Diseases Information Center (GARD), but also support the public health of the rare disease community.

Rare disease-based scientific annotation knowledge graph

Rare diseases (RDs) are naturally associated with a low prevalence rate, which raises a big challenge due to there being less data available for supporting preclinical and clinical studies. There has been a vast improvement in our understanding of RD, largely owing to advanced big data analytic approaches in genetics/genomics. Consequently, a large volume of RD-related publications has been accumulated in recent years, which offers opportunities to utilize these publications for accessing the full spectrum of the scientific research and supporting further investigation in RD. In this study, we systematically analyzed, semantically annotated, and scientifically categorized RD-related PubMed articles, and integrated those semantic annotations in a knowledge graph (KG), which is hosted in Neo4j based on a predefined data model. With the successful demonstration of scientific contribution in RD via the case studies performed by exploring this KG, we propose to extend the current effort by expanding more RD-related publications and more other types of resources as a next step.

Lipidomics and Redox Lipidomics Indicate Early Stage Alcohol-Induced Liver Damage

Alcoholic fatty liver disease (AFLD) is characterized by lipid accumulation and inflammation and can progress to cirrhosis and cancer in the liver. AFLD diagnosis currently relies on histological analysis of liver biopsies. Early detection permits interventions that would prevent progression to cirrhosis or later stages of the disease. Herein, we have conducted the first comprehensive time-course study of lipids using novel state-of-the art lipidomics methods in plasma and liver in the early stages of a mouse model of AFLD, i.e., Lieber-DeCarli diet model. In ethanol-treated mice, changes in liver tissue included up-regulation of triglycerides (TGs) and oxidized TGs and down-regulation of phosphatidylcholine, lysophosphatidylcholine, and 20-22-carbon-containing lipid-mediator precursors. An increase in oxidized TGs preceded histological signs of early AFLD, i.e., steatosis, with these changes observed in both the liver and plasma. The major lipid classes dysregulated by ethanol play important roles in hepatic inflammation, steatosis, and oxidative damage. Conclusion: Alcohol consumption alters the liver lipidome before overt histological markers of early AFLD. This introduces the exciting possibility that specific lipids may serve as earlier biomarkers of AFLD than those currently being used.

Identification of a Putative Enhancer RNA for EGFR in Hyper-Accessible Regions in Esophageal Squamous Cell Carcinoma Cells by Analysis of Chromatin Accessibility Landscapes

Abnormal genetic and epigenetic modifications play a key role in esophageal cancer. By Assay for Transposase-Accessible Chromatin by sequencing (ATAC-seq), this study compared chromatin accessibility landscapes among two esophageal squamous cell carcinoma (ESCC) cell lines, KYSE-30 and KYSE-150, and a non-cancerous esophageal epithelial cell line, HET-1A. Data showed that hyper-accessible regions in ESCC cells contained genes related with cancer hallmarks, such as epidermal growth factor receptor (EGFR). Multi-omics analysis and digital-droplet PCR results demonstrated that several non-coding RNAs in EGFR upstream were upregulated in ESCC cells. Among them, one appeared to act as an enhancer RNA responsible for EGFR overexpression. Further motif analysis and pharmacological data suggested that AP-1 family transcription factors were able to bind the hyper-accessible regions and thus to regulate cancer cell proliferation and migration. This study discovered a putative enhancer RNA for EGFR gene and the reliance of ESCC on AP-1 transcription factor.

Validating ADME QSAR Models Using Marketed Drugs

Problems with drug ADME are responsible for many clinical failures. By understanding the ADME properties of marketed drugs and modeling how chemical structure contributes to these inherent properties, we can help new projects reduce their risk profiles. Kinetic aqueous solubility, the parallel artificial membrane permeability assay (PAMPA), and rat liver microsomal stability constitute the Tier I ADME assays at the National Center for Advancing Translational Sciences (NCATS). Using recent data generated from in-house lead optimization Tier I studies, we update quantitative structure-activity relationship (QSAR) models for these three endpoints and validate in silico performance against a set of marketed drugs (balanced accuracies range between 71% and 85%). Improved models and experimental datasets are of direct relevance to drug discovery projects and, together with the prediction services that have been made publicly available at the ADME@NCATS web portal (https://opendata.ncats.nih.gov/adme/), provide important tools for the drug discovery community. The results are discussed in light of our previously reported ADME models and state-of-the-art models from scientific literature.Graphical Abstract[Figure: see text].

Network analyses in microbiome based on high-throughput multi-omics data

Together with various hosts and environments, ubiquitous microbes interact closely with each other forming an intertwined system or community. Of interest, shifts of the relationships between microbes and their hosts or environments are associated with critical diseases and ecological changes. While advances in high-throughput Omics technologies offer a great opportunity for understanding the structures and functions of microbiome, it is still challenging to analyse and interpret the omics data. Specifically, the heterogeneity and diversity of microbial communities, compounded with the large size of the datasets, impose a tremendous challenge to mechanistically elucidate the complex communities. Fortunately, network analyses provide an efficient way to tackle this problem, and several network approaches have been proposed to improve this understanding recently. Here, we systemically illustrate these network theories that have been used in biological and biomedical research. Then, we review existing network modelling methods of microbial studies at multiple layers from metagenomics to metabolomics and further to multi-omics. Lastly, we discuss the limitations of present studies and provide a perspective for further directions in support of the understanding of microbial communities.

TCRD and Pharos 2021: mining the human proteome for disease biology

In 2014, the National Institutes of Health (NIH) initiated the Illuminating the Druggable Genome (IDG) program to identify and improve our understanding of poorly characterized proteins that can potentially be modulated using small molecules or biologics. Two resources produced from these efforts are: The Target Central Resource Database (TCRD) (http://juniper.health.unm.edu/tcrd/) and Pharos (https://pharos.nih.gov/), a web interface to browse the TCRD. The ultimate goal of these resources is to highlight and facilitate research into currently understudied proteins, by aggregating a multitude of data sources, and ranking targets based on the amount of data available, and presenting data in machine learning ready format. Since the 2017 release, both TCRD and Pharos have produced two major releases, which have incorporated or expanded an additional 25 data sources. Recently incorporated data types include human and viral-human protein-protein interactions, protein-disease and protein-phenotype associations, and drug-induced gene signatures, among others. These aggregated data have enabled us to generate new visualizations and content sections in Pharos, in order to empower users to find new areas of study in the druggable genome.

Correction to: The International Conference on Intelligent Biology and Medicine (ICIBM) 2019: bioinformatics methods and applications for human diseases

After publication of this supplement article [1], it is requested the grant ID in the Funding section should be corrected from NSF grant IIS-7811367 to NSF grant IIS-1902617. Therefore, the correct 'Funding' section in this article should read: We thank the National Science Foundation (NSF grant IIS-1902617) for the financial support of ICIBM 2019. This article has not received sponsorship for publication.

The International Conference on Intelligent Biology and Medicine 2019 (ICIBM 2019): computational methods and applications in medical genomics

In this editorial, we briefly summarized the International Conference on Intelligent Biology and Medicine 2019 (ICIBM 2019) that was held on June 9-11, 2019 at Columbus, Ohio, USA. We further introduced the 19 research articles included in this supplement issue, covering four major areas, namely computational method development, genomics analysis, network-based analysis and biomarker prediction. The selected papers perform cutting edge computational research applied to a broad range of human diseases such as cancer, neural degenerative and chronic inflammatory disease. They also proposed solutions for fundamental medical genomics problems range from basic data processing and quality control to functional interpretation, biomarker and drug prediction, and database releasing.

The International Conference on Intelligent Biology and Medicine 2019: computational methods for drug interactions

In this editorial, we briefly summarize the International Conference on Intelligent Biology and Medicine 2019 (ICIBM 2019) that was held on June 9-11, 2019 at Columbus, Ohio, USA. Then, we introduce the two research articles included in this supplement issue. These two research articles were selected after careful review of 105 articles that were submitted to the conference, and cover topics on deep learning for drug-target interaction prediction and data mining and visualization of high-order drug-drug interactions.

The International Conference on Intelligent Biology and Medicine 2019 (ICIBM 2019): conference summary and innovations in genomics

The goal of this editorial is to summarize the 2019 International Conference on Intelligent Biology and Medicine (ICIBM 2019) conference that took place on June 9–11, 2019 in The Ohio State University, Columbus, OH, and to provide an introductory summary of the seven articles presented in this supplement issue. ICIBM 2019 hosted four keynote speakers, four eminent scholar speakers, five tutorials and workshops, twelve concurrent sessions and a poster session, totaling 23 posters, spanning state-of-the-art developments in bioinformatics, genomics, next-generation sequencing (NGS) analysis, scientific databases, cancer and medical genomics, and computational drug discovery. A total of 105 original manuscripts were submitted to ICIBM 2019, and after careful review, seven were selected for this supplement issue. These articles cover methods and applications for functional annotations of miRNA targeting, clonal evolution of bacterial cells, gene co-expression networks that describe a given phenotype, functional binding site analysis of RNA-binding proteins, normalization of genome architecture mapping data, sample predictions based on multiple NGS data types, and prediction of an individual’s genetic admixture given exonic single nucleotide polymorphisms data.

The International Conference on Intelligent Biology and Medicine (ICIBM) 2019: bioinformatics methods and applications for human diseases

Between June 9-11, 2019, the International Conference on Intelligent Biology and Medicine (ICIBM 2019) was held in Columbus, Ohio, USA. The conference included 12 scientific sessions, five tutorials or workshops, one poster session, four keynote talks and four eminent scholar talks that covered a wide range of topics in bioinformatics, medical informatics, systems biology and intelligent computing. Here, we describe 13 high quality research articles selected for publishing in BMC Bioinformatics.

Challenges in proteogenomics: a comparison of analysis methods with the case study of the DREAM proteogenomics sub-challenge

Background

Proteomic measurements, which closely reflect phenotypes, provide insights into gene expression regulations and mechanisms underlying altered phenotypes. Further, integration of data on proteome and transcriptome levels can validate gene signatures associated with a phenotype. However, proteomic data is not as abundant as genomic data, and it is thus beneficial to use genomic features to predict protein abundances when matching proteomic samples or measurements within samples are lacking.

Results

We evaluate and compare four data-driven models for prediction of proteomic data from mRNA measured in breast and ovarian cancers using the 2017 DREAM Proteogenomics Challenge data. Our results show that Bayesian network, random forests, LASSO, and fuzzy logic approaches can predict protein abundance levels with median ground truth-predicted correlation values between 0.2 and 0.5. However, the most accurately predicted proteins differ considerably between approaches.

Conclusions

In addition to benchmarking aforementioned machine learning approaches for predicting protein levels from transcript levels, we discuss challenges and potential solutions in state-of-the-art proteogenomic analyses.

The metaRbolomics Toolbox in Bioconductor and beyond

Metabolomics aims to measure and characterise the complex composition of metabolites in a biological system. Metabolomics studies involve sophisticated analytical techniques such as mass spectrometry and nuclear magnetic resonance spectroscopy, and generate large amounts of high-dimensional and complex experimental data. Open source processing and analysis tools are of major interest in light of innovative, open and reproducible science. The scientific community has developed a wide range of open source software, providing freely available advanced processing and analysis approaches. The programming and statistics environment R has emerged as one of the most popular environments to process and analyse Metabolomics datasets. A major benefit of such an environment is the possibility of connecting different tools into more complex workflows. Combining reusable data processing R scripts with the experimental data thus allows for open, reproducible research. This review provides an extensive overview of existing packages in R for different steps in a typical computational metabolomics workflow, including data processing, biostatistics, metabolite annotation and identification, and biochemical network and pathway analysis. Multifunctional workflows, possible user interfaces and integration into workflow management systems are also reviewed. In total, this review summarises more than two hundred metabolomics specific packages primarily available on CRAN, Bioconductor and GitHub.

Metabolomics Analytics Workflow for Epidemiological Research: Perspectives from the Consortium of Metabolomics Studies (COMETS)

The application of metabolomics technology to epidemiological studies is emerging as a new approach to elucidate disease etiology and for biomarker discovery. However, analysis of metabolomics data is complex and there is an urgent need for the standardization of analysis workflow and reporting of study findings. To inform the development of such guidelines, we conducted a survey of 47 cohort representatives from the Consortium of Metabolomics Studies (COMETS) to gain insights into the current strategies and procedures used for analyzing metabolomics data in epidemiological studies worldwide. The results indicated a variety of applied analytical strategies, from biospecimen and data pre-processing and quality control to statistical analysis and reporting of study findings. These strategies included methods commonly used within the metabolomics community and applied in epidemiological research, as well as novel approaches to pre-processing pipelines and data analysis. To help with these discrepancies, we propose use of open-source initiatives such as the online web-based tool COMETS Analytics, which includes helpful tools to guide analytical workflow and the standardized reporting of findings from metabolomics analyses within epidemiological studies. Ultimately, this will improve the quality of statistical analyses, research findings, and study reproducibility.

Abstract 664: Microbial and inflammatory response to electronic cigarette and cigarette use

Background: As electronic cigarette (e-cig) use increases in the US, investigation of their effects are critical. Alteration of the lung microbiome, oral microbiome, and inflammation are well established effects of cigarettes; however the effects of e-cigs are yet unknown. Individuals with smoking-related lung disease have alterations in their lung microbiome compared to healthy individuals and their lung microbiomes appear more similar to their oral microbiomes when compared to healthy individuals. To our knowledge, only one study has examined smoking tobacco’s concurrent effects in the oral and lung microbiome and none have examined e-cig use. Further, none have studied e-cigs’ effect on the lung microbiome and inflammation. We hypothesized that e-cig use would affect the lung microbiome, and that the effects are different from smokers and never-smokers; alteration of the lung microbiome will also affect inflammatory gene expression in the lungs.

Methods: A cross-sectional study of bronchoscopy with bronchoalveolar lavage (BAL) of 10 never-smokers, 8 cigarette smokers, and 10 e-cig users was conducted. RNA was extracted from BAL samples for total transcriptome RNA-seq analysis, allowing measurement of the microbiome and human gene expression. Differences in the microbiome by smoking status were determined by the Kruskal-Wallis test. Pairwise Wilcoxon rank sum tests with Holm correction was used. Effect size (fold change >1.5) and adjusted P-value cutoffs (<0.05) were used to identify microbes of potential interest. The limma-voom package in R was used to determine associations with human gene expression.

Results: We identified 53 differentially-abundant bacterial species in BAL samples by smoking group. Among them, the majority were less abundant in the lung of smokers and ~20 are normally found in the oral microbiome. While there were significant differences in differentially-abundant microbes between e-cig users and smokers and between smokers and never-smokers, the microbiome of e-cig users did not differ from that of never-smokers. In preliminary analyses of gene expression, there were 2,400 differentially-expressed human genes among the three groups, of which 58 are inflammatory pathway genes.

Conclusion: The majority of differentially-abundant microbes observed by smoking group are largely due to smokers. The microbiome of e-cig users is more similar to that of never-smokers. Interestingly, nearly half of microbes that are altered in the lung microbiome due to smoking use are bacterial species normally found in the oral microbiome. These findings suggest that the alterations in the oral microbiome associated with smoking cigarettes may also be reflected in the lung microbiome.

Citation Format: Kevin L. Ying, Min-Ae Song, Daniel Y. Weng, Quentin A. Nickerson, Joseph P. McElroy, Theodore M. Brasky, Mark D. Wewers, Ewy Mathé, Jo L. Freudenheim, Peter G. Shields. Microbial and inflammatory response to electronic cigarette and cigarette use [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 664.

Integration of Metabolomic and Other Omics Data in Population-Based Study Designs: An Epidemiological Perspective

It is not controversial that study design considerations and challenges must be addressed when investigating the linkage between single omic measurements and human phenotypes. It follows that such considerations are just as critical, if not more so, in the context of multi-omic studies. In this review, we discuss (1) epidemiologic principles of study design, including selection of biospecimen source(s) and the implications of the timing of sample collection, in the context of a multi-omic investigation, and (2) the strengths and limitations of various techniques of data integration across multi-omic data types that may arise in population-based studies utilizing metabolomic data.

The Consortium of Metabolomics Studies (COMETS): Metabolomics in 47 Prospective Cohort Studies

The Consortium of Metabolomics Studies (COMETS) was established in 2014 to facilitate large-scale collaborative research on the human metabolome and its relationship with disease etiology, diagnosis, and prognosis. COMETS comprises 47 cohorts from Asia, Europe, North America, and South America that together include more than 136,000 participants with blood metabolomics data on samples collected from 1985 to 2017. Metabolomics data were provided by 17 different platforms, with the most frequently used labs being Metabolon, Inc. (14 cohorts), the Broad Institute (15 cohorts), and Nightingale Health (11 cohorts). Participants have been followed for a median of 23 years for health outcomes including death, cancer, cardiovascular disease, diabetes, and others; many of the studies are ongoing. Available exposure-related data include common clinical measurements and behavioral factors, as well as genome-wide genotype data. Two feasibility studies were conducted to evaluate the comparability of metabolomics platforms used by COMETS cohorts. The first study showed that the overlap between any 2 different laboratories ranged from 6 to 121 metabolites at 5 leading laboratories. The second study showed that the median Spearman correlation comparing 111 overlapping metabolites captured by Metabolon and the Broad Institute was 0.79 (interquartile range, 0.56-0.89).

Integration of Metabolomics and Transcriptomics to Identify Gene-Metabolite Relationships Specific to Phenotype

Metabolomics plays an increasingly large role in translational research, with metabolomics data being generated in large cohorts, alongside other omics data such as gene expression. With this in mind, we provide a review of current approaches that integrate metabolomic and transcriptomic data. Furthermore, we provide a detailed framework for integrating metabolomic and transcriptomic data using a two-step approach: (1) numerical integration of gene and metabolite levels to identify phenotype (e.g., cancer)-specific gene-metabolite relationships using IntLIM and (2) knowledge-based integration, using pathway overrepresentation analysis through RaMP, a comprehensive database of biological pathways. Each step makes use of publicly available R packages ( https://github.com/mathelab/IntLIM and https://github.com/mathelab/RaMP-DB ), and provides a user-friendly web interface for analysis. These interfaces can be run locally through the package or can be accessed through our servers ( https://intlim.bmi.osumc.edu and https://ramp-db.bmi.osumc.edu ). The goal of this chapter is to provide step-by-step instructions on how to install the software and use the commands within the R framework, without the user interface (which is slower than running the commands through command line). Both packages are in continuous development so please refer to the GitHub sites to check for updates.

The Omics Revolution Continues: The Maturation of High-Throughput Biological Data Sources

OBJECTIVE

 The aim is to provide a comprehensive review of state-of-the art omics approaches, including proteomics, metabolomics, cell-free DNA, and patient cohort matching algorithms in precision oncology.

METHODS

 In the past several years, the cancer informatics revolution has been the beneficiary of a data explosion. Different complementary omics technologies have begun coming into their own to provide a more nuanced view of the patient-tumor interaction beyond that of DNA alterations. A combined approach is beneficial to the patient as nearly all new cancer therapeutics are designed with an omics biomarker in mind. Proteomics and metabolomics provide us with a means of assaying in real-time the response of the tumor to treatment. Circulating cell-free DNA may allow us to better understand tumor heterogeneity and interactions with the host genome.

RESULTS

 Integration of increasingly available omics data increases our ability to segment patients into smaller and smaller cohorts, thereby prompting a shift in our thinking about how to use these omics data. With large repositories of patient omics-outcomes data being generated, patient cohort matching algorithms have become a dominant player.

CONCLUSIONS

 The continued promise of precision oncology is to select patients who are most likely to benefit from treatment and to avoid toxicity for those who will not. The increased public availability of omics and outcomes data in patients, along with improved computational methods and resources, are making precision oncology a reality.

Abstract 3088: Deciphering components of mutational signatures arising from carcinogen co-exposures: A genome-scale experimental approach

Pan-cancer analyses of tumor genomes reveal mutational signatures characteristic of particular etiologic factors. The signatures are extracted mathematically from mixed patterns typically observed by tumor sequencing. However, the components of signatures originating from complex carcinogen mixtures have not been studied in detail. Lung, head and neck and liver tumors of tobacco smokers exhibit COSMIC signature 4 marked by predominant G>N mutations, involving mainly G:C>T:A transversions with transcription strand bias, consistent with the mutagenic effects of benzo[a]pyrene (B[a]P). Additionally, A>N mutations (strand-biased A:T>T:A transversions and A:T>G:C transitions) are also prominently present, yet their origins are less understood. By using exposure-coupled clonal immortalization of human and mouse primary cells and deep sequencing, we were able to dissect ‘clean' mutational signatures of tobacco smoke carcinogens B[a]P and glycidamide (GA), a key reactive metabolite of acrylamide (ACR). Whole-genome sequencing of multiple clones derived from primary B[a]P-treated human mammary epithelial cells identified a robust mutational signature marked by strand-biased G>N mutations and increased GG>TT dinucleotides, while no apparent enrichment of A:T>T:A mutations was observed. Next, in ACR and GA-treated primary mouse embryonic fibroblasts, we established by the LC-MS/MS DNA adduct analysis that ACR exerts its mutagenic effects exclusively via GA. We then extracted from 15 treated clones the exome-scale mutational signature of GA, marked by predominant A:T>T:A transversions followed by A:T>G:C transitions and G:C>T:A transversions, all showing transcription strand bias. Similarity analysis involving known primary-cancer and experimental mutational signatures indicated that the GA mutational signature was novel. A more in-depth comparison with mutation patterns from lung adenocarcinomas of heavy smokers revealed that the GA signature, including its strand bias features, matched closely with and may thus account for the A>N mutation component of the tobacco smoking-derived signature 4. Thus, mutational signatures generated in controlled experimental settings may explain particular sub-features of cancer signatures arising from co-exposures to multiple carcinogens. Furthermore, the use of innovative in vitro systems, characterized by biological barrier bypass to mimic early steps of cell transformation, can provide revealing insights into the molecular links between mutagenesis and carcinogenesis. Funding: INCa-INSERM Plan Cancer 2015; NIH/NIEHS 1R03ES025023-01A1

Citation Format: Manuraj Pandey, Maria Zhivagui, Mona I. Churchwell, Alvin W. Ng, Liacine Bouaoun, Vincent Cahais, Martha R. Stampfer, Magali Olivier, Zdenko Herceg, Ewy Mathé, Steven G. Rozen, Frederick A. Beland, Michael Korenjak, Jiri Zavadil. Deciphering components of mutational signatures arising from carcinogen co-exposures: A genome-scale experimental approach [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3088.

Abstract 3529: Modulation of MDM2 alters the metabolomic programming of dedifferentiated liposarcoma and its sensitivity to cholesterol inhibition

BACKGROUND: Dedifferentiated liposarcoma (DDLPS) is a highly morbid, mesenchymal tumor characterized by amplification of the 12q chromosomal loci. Although DDLPS are lipid tumors, they produce little lipid. Life expectancy is typically less than 15 months in the advanced setting and standard treatments remain highly toxic. Superior treatment options are clearly needed. Amplification of the MDM2 oncogene that resides in these loci is observed in 100% of all DDLPS; however, to variable levels. Through a negative feedback loop, MDM2 inhibits the tumor suppressive functions of p53 that halts cell growth after cellular stress. Previously, we have reported that MDM2 levels correspond with cellular growth and drug metabolism, but little is known about the effects of MDM2 alterations on global metabolomic profiles. These profiles will help us pinpoint dysregulated pathways that explain, at least partially, the functional effects associated with MDM2 amplification. METHODS: Six DDLPS cell lines were brought directly into culture from patients. MDM2 levels were determined via Western blot and RNA-sequencing. Metabolomics data was generated using the Metabolon platform. Cell viability assays were performed in ZOOM IncuCyte or measured by XTT. Atorvastatin was used to inhibit cholesterol synthesis. MDM2 levels were altered using SAR405838 to raise MDM2 levels and MI-192 to lower MDM2 levels. Synergy was calculated via Chou-Talalay method to determine the combination index (CI) using Compusyn software. RESULTS: MDM2 levels are inversely correlated with metabolites in the lipid and cholesterol pathway (Fisher's, p < 0.001). The lipid metabolism pathway was also the top deregulated pathway (pathway enrichment p-value = 0.03) in transcriptionally profiled DDLPS cell lines treated with MDM2 elevating agent SAR405838. MDM2 low DDLPS cell lines were exquisitely sensitive to HMGA-CoA reductase inhibitors in the low micromolar range. Lipid metabolite profiling of MDM2 low versus high cell lines treated with atorvastatin demonstrated that twice as many lipid metabolites were altered in MDM2 low versus high cells (Chi-square, p < 0.001). MDM2 levels by RNA-seq demonstrated significant correlation between MDM2 gene expression and atorvastatin IC50 doses (r=0.963). MDM2 modulation by use of SAR405838 and MI-192 respectively in combination with atorvastatin displayed antagonism (average CI = 1.5) and synergy (average CI = 0.63), respectively. CONCLUSION: Modulation of MDM2 alters cholesterol metabolism in DDLPS and may serve as druggable target.

Citation Format: Bryce Demoret, Andrew Patt, John Hays, Ewy Mathé, James L. Chen. Modulation of MDM2 alters the metabolomic programming of dedifferentiated liposarcoma and its sensitivity to cholesterol inhibition [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3529.

Abstract 1231: Using oral and lung microbiome to assess microbial dysbiosis and inflammatory response to electronic cigarettes and to cigarettes

Background: Alteration of the oral microbiome (microbial dysbiosis) with cigarette smoking is well established. However, the effect of electronic cigarettes (e-cigs) use on the oral microbiome is unknown, although there are emerging data that e-cigs induce microbial changes similar to smoking. In smoking-related diseases, such as chronic obstructive pulmonary disease, there are changes in the oral microbiome and in the expression of genes involved in inflammatory pathways. Similar to the oral microbiome, it is feasible that smoking tobacco and e-cig use could also affect the lung microbiome. To the best of our knowledge, there is only one published study investigating smoking tobacco effects on the oral and lung microbiome. No published studies have evaluated concurrent effects of e-cigs in the oral and lung microbiome.

Aims: We hypothesize that microbial dysbiosis and expression of inflammatory cytokines in the oral cavity and lung will differ between smokers and nonsmokers, and that e-cig users will have microbial dysbiosis more similar to smokers. To accomplish this, we propose 1) to examine the association of oral and lung microbiome in nonsmokers, smokers and e-cig users, 2) to determine if the oral microbiome and the lung microbiome differ among these groups, and 3) to determine correlation of the microbiota with host expression of inflammation-related genes.

Methods: A cross-sectional study using bronchoscopy and oral rinse collection of 10 never-smokers, 8 cigarette smokers, and 10 e-cig users was conducted. For each study participant, RNA was extracted from saliva and bronchoalveolar lavage (BAL) samples for total transcriptome analysis using RNA-seq; facilitating this approach allows measurement of bacterial communities and human inflammatory cytokine expression in the same assay. To determine microbial dysbiosis by smoking status, the Mann Whitney U-test and Kruskal-Wallis H-test were used with Bonferroni correction for multiple comparisons. Both effect size (fold change >1.5) and adjusted p-value cutoffs (<0.05) were used to identify statistical significance.

Results: In preliminary analyses we identified 2,257 bacterial strains in saliva samples and 1592 in BAL samples. We found a lack of concordance of highly abundant bacteria in the oral cavity and lungs. The top twenty expressed human genes were associated with RNA splicing, RNA elongation and miRNAs. Comparisons of microbial dysbiosis by smoking status are currently under way.

Conclusion: The composition of the microbiome for saliva is different from that of BAL. Comparison of the metatranscriptome and transcriptome between the lung and oral cavity, as well as between smokers, nonsmokers and e-cigarette users, will allow us to observe how e-cig use compares with cigarette smoking and never smoking in terms of microbial dysbiosis and inflammatory cytokines.

Citation Format: Kevin L. Ying, Min-Ae Song, Daniel Y. Weng, Quentin A. Nickerson, Joseph P. McElroy, David Frankhouser, Pearlly S. Yan, Ralf Bundschuh, Theodore M. Brasky, Mark D. Wewers, Ewy Mathé, Jo L. Freudenheim, Peter G. Shields. Using oral and lung microbiome to assess microbial dysbiosis and inflammatory response to electronic cigarettes and to cigarettes [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1231.

Matchmaking in Bioinformatics

Ever return from a meeting feeling elated by all those exciting talks, yet unsure how all those presented glamorous and/or exciting tools can be useful in your research?  Or do you have a great piece of software you want to share, yet only a handful of people visited your poster? We have all been there, and that is why we organized the Matchmaking for Computational and Experimental Biologists Session at the latest ISCB/GLBIO’2017 meeting in Chicago (May 15-17, 2017). The session exemplifies a novel approach, mimicking “matchmaking”, to encouraging communication, making connections and fostering collaborations between computational and non-computational biologists. More specifically, the session facilitates face-to-face communication between researchers with similar or differing research interests, which we feel are critical for promoting productive discussions and collaborations.  To accomplish this, three short scheduled talks were delivered, focusing on RNA-seq, integration of clinical and genomic data, and chromatin accessibility analyses.  Next, small-table developer-led discussions, modeled after speed-dating, enabled each developer (including the speakers) to introduce a specific tool and to engage potential users or other developers around the table.  Notably, we asked the audience whether any other tool developers would want to showcase their tool and we thus added four developers as moderators of these small-table discussions.  Given the positive feedback from the tool developers, we feel that this type of session is an effective approach for promoting valuable scientific discussion, and is particularly helpful in the context of conferences where the number of participants and activities could hamper such interactions.

Abstract 246: Assessing microbial dysbiosis of electronic cigarettes and cigarette smokers using oral and lung microbiome

The link between smoking tobacco and changes in the oral microbiome in response to tobacco smoking are well established. It is not known if there are changes in response to electronic cigarettes (e-cig). These changes in the microbiome are associated with increased numbers of disease causing pathogens. Currently there are no published studies that have investigated the relationship of smoking tobacco on both the oral and lung microbiome. There is insufficient evidence showing whether changes in oral cavity and lung microbiome are also seen in e-cig users. We will study the oral cavity and lung of non-smokers, smokers and e-cig users to examine concordance between oral cavity and the lungs as well as comparing the three groups, examining the microbiomes and expression of inflammatory markers. We hypothesize that microbial dysbiosis and expression of inflammatory cytokines will differ for smokers and non-smokers; and that e-cig users will have microbial dysbiosis similar to cigarette smokers. A cross-sectional study is being conducted on three groups, 1) never-smokers, 2) cigarette smokers, and 3) e-cig users. For each study participant, saliva and bronchoalveolar lavage (BAL) are being collected to measure microbiome. RNA is extracted from saliva and BAL samples for total transcriptome analysis using RNA-seq. This analysis will detect human and bacterial reads thereby allowing observations of bacterial communities as well as human inflammatory cytokine response to bacterial presence. 85% to 98% of BAL sample reads aligned to the human genome compared to less than 50% from saliva samples. The alignment results allow us to deduce that the majority of reads from BAL samples are human and that the majority of the reads in saliva samples are bacterial.

Preliminary results show detection of human RNA expression and of bacterial reads are present in both saliva and BAL samples. More samples are being processed and the comparison of BAL and saliva samples between the three groups will be discussed.

Citation Format: Kevin Ying, Min-Ae Song, Daniel Y. Weng, Quentin Nickerson, David Frankhouser, Pearlly S. Yan, Ralf Bundschuh, Mark D. Wewers, Ewy Mathé, Jo L. Freudenheim, Peter G. Shields. Assessing microbial dysbiosis of electronic cigarettes and cigarette smokers using oral and lung microbiome [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 246. doi:10.1158/1538-7445.AM2017-246

B cell super-enhancers and regulatory clusters recruit AID tumorigenic activity

The antibody gene mutator activation-induced cytidine deaminase (AID) promiscuously damages oncogenes, leading to chromosomal translocations and tumorigenesis. Why nonimmunoglobulin loci are susceptible to AID activity is unknown. Here, we study AID-mediated lesions in the context of nuclear architecture and the B cell regulome. We show that AID targets are not randomly distributed across the genome but are predominantly grouped within super-enhancers and regulatory clusters. Unexpectedly, in these domains, AID deaminates active promoters and eRNA(+) enhancers interconnected in some instances over megabases of linear chromatin. Using genome editing, we demonstrate that 3D-linked targets cooperate to recruit AID-mediated breaks. Furthermore, a comparison of hypermutation in mouse B cells, AID-induced kataegis in human lymphomas, and translocations in MEFs reveals that AID damages different genes in different cell types. Yet, in all cases, the targets are predominantly associated with topological complex, highly transcribed super-enhancers, demonstrating that these compartments are key mediators of AID recruitment.

Abstract A47: Epigenetic regulation of microRNAs in early stage lung adenocarcinoma

Lung cancer is the most commonly diagnosed cancer and a leading cause of cancer-associated deaths worldwide. It is imperative to understand the molecular mechanisms underlying early lung cancer development so that better diagnostic and treatment options can be identified. We conducted a multi-platform genomic screen of stage I/II lung adenocarcinoma and non-tumor adjacent tissue excised from patients recruited into our ongoing NCI-MD Case-Control Study. MicroRNAs (miRNAs) are 20-22 nucleotide non-coding RNAs that regulate mRNA transcription and translation through sequence-specific base pairing. miRNAs are transcribed from large pri-miRNAs as independent transcripts or within introns of protein-coding genes. Although it is unclear whether they possess independent promoter regions, several miRNAs are adjacent to CpG islands. CpG island methylation is a well-described mechanism by which genes and miRNAs with tumor-suppressive function are silenced in cancer. Such regulation could have important implications for progression of early stage lung cancer. Methods: Forty paired tumor and adjacent non-tumor tissues were characterized through methylation profiling using Illumina Infinium HumanMethylation27 BeadChip assays, miRNA expression profiling using Nanostring nCounter miRNA Expression Assay and mRNA expression profiling using Illumina HumanRef-8 v3 Expression Beadchip arrays. Consistency of methylation profiling was compared across three other cohorts of lung adenocarcinoma: TCGA (The Cancer Genome Atlas) lung adenocarcinoma (LUAD) consisting of 23 matched-tissue pairs downloaded from the TCGA data portal; GSE32867, comprising 60 matched-tissue pairs published by Selamat et al. (Genome Res. 2012 Jul;22(7)) and downloaded from GEO; and a lung adenocarcinoma cohort from Japan consisting of 76 matched-tissue pairs. Additionally, in order to functionally validate miRNA-associated methylation we exposed A549 lung cancer cell line to a chromatin demethylation protocol of 5-aza/TSA. Results: We found seven miRNA loci hypermethylated in tumors. Four of these are located within homeobox gene clusters, miR-196b (HOXA9), miR-615 (HOXC5), miR-10a (HOXB4), and miR-10b (HOXD4). Two others are located adjacent to genes with significant changes in methylation, miR-638 (DNM2) and miR-639 (GPSN2). Increased methylation of miR-34b/c (BTG4) in lung tumors was consistent with prior studies. A comparative analysis of methylation at these loci showed a high concordance between our study and methylation profiling in three other cohorts of lung adenocarcinoma. Through microRNA expression profiling, we determined that miR-520e is the top downregulated miRNA in tumors. Interestingly, miR-520e belongs to a large microRNA cluster in 19q13.42 (miR-515 family) flanked by a CpG island. Functional validation of putative tumor-suppressor microRNAs is ongoing. We have confirmed that miR-615 and miR-520e are methylated and become demethylated and reactivated by exposure to a combination of 5-aza and TSA in-vitro. Conclusions: CpG sites associated with miRNAs are methylated in early stage lung adenocarcinoma across independent cohorts of lung cancer. Therefore, tumor suppressive miRNAs maybe epigenetically controlled through CpG island methylation in early stage lung adenocarcinoma. Analysis of their association with patient outcome is forthcoming.

Citation Format: Ana I. Robles, Hirokazu Okayama, Saikartik A. Kumar, Zaneta Franklin, Elise D. Bowman, Daniel Edelman, Holly Stevenson, David Petersen, Ewy Mathé, Yae Kanai, Takashi Kohno, Naoto Tsuchiya, Paul Meltzer, Jun Yokota, Curtis C. Harris. Epigenetic regulation of microRNAs in early stage lung adenocarcinoma. [abstract]. In: Proceedings of the Third AACR International Conference on Frontiers in Basic Cancer Research; Sep 18-22, 2013; National Harbor, MD. Philadelphia (PA): AACR; Cancer Res 2013;73(19 Suppl):Abstract nr A47.

Inflammation regulates microRNA expression in cooperation with p53 and nitric oxide

microRNA (miRNA) are small non-coding RNA targeting mRNAs leading to their instability and diminished translation. Altered expression of miRNA is associated with cancer. Inflammation and nitric oxide modulates the development of lymphomas in p53 knockout mice and there exists a negative feedback loop between p53 and NOS2. Using a genetic strategy, we tested the hypothesis that inflammation-induced oxidative and nitrosative stress modulates miRNA expression in mouse model deficient in either p53 or NOS2. Mice treated with Corynebacterium parvum (C. parvum), to induce inflammation, clearly separated from controls by their miRNA profiles in wild-type, p53- and NOS2-knockout genetic backgrounds. C. parvum-induced inflammation significantly (p < 0.005) increased miR-21, miR-29b and miR-34a/b/c and decreased (p < 0.005) mir-29c and mir-181a/c expression in the spleen of C57BL mice. However, p53-knockout C57BL mice did not show a significant increase in the mir-34b/c or a decrease in mir-29c expression following C. parvum-induced inflammation. Expression of mir-21, mir-29b and mir-181a was independent of p53-status. NOS2-knockout C57BL mice showed a significant increase in miR-21 and miR-34a/b/c and decrease in miR-181a similar to the wild-type (WT) mice following C. parvum-induced inflammation. However, in contrast to the WT mice, miR-29b/c expression was not affected following C. parvum-induced inflammation in NOS2 knockout mice. N-acetyl cysteine, an anti-oxidant, reduced the expression of miR-21 and miR-29b in C. parvum-treated WT mice (p < 0.005) as compared with control C. parvum-treated mice. These data are consistent with the hypothesis that inflammation modulates miRNA expression in vivo and the alteration in specific miRNA under an inflammatory microenvironment, can be influenced by p53 (miR-34b/c) and NO(•) (29b/c).