From Chromatin Structure to Signal Transduction,
through Transcription and RNA Stability 

DAY 1

MONDAY, OCTOBER 17

7:30-8:00 Conference Registration and Morning Coffee

8:00-8:25 Regulatory Networks Overview and Opening Remarks
Thomas Werner, Ph.D., CEO & CSO, Genomatix Software GmbH
Regulatory networks consist of several interconnected layers ranging from signal transduction pathways involving kinase cascades to transcription control and posttranscriptional regulation of RNA pools, e.g., by microRNAs. However, many intermediate steps in these series of events directly influence other layers of gene regulation often linking simultaneously to more than one layer. The tight interconnection of signaling, transcription control and epigenetic changes becomes evident on the example of breast cancer related regulatory networks. It is essential to include all layers of gene regulation as functional gene-gene dependencies and feedback loops often involve different levels including protein and DNA modification as well as transcriptional changes. This wholesome approach is required to reveal biologically important interdependencies such as functional connections of tumor suppressor genes.

Chairperson: Thomas Werner, Ph.D.
8:25-8:55 Chromatin Immunoprecipitation
Kevin Struhl, Ph.D., David Wesley Gaiser Professor, Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School
Chromatin immunoprecipitation is a method that allows one to identify the specific locations in the genome where proteins of interest are bound in living cells. When combined with tiled microarrays representing whole genomes, it is possible to identify target sites of transcriptional regulatory proteins and specific features of chromatin on a genome-wide basis in an unbiased manner. These methodologies have provided new insight into the mechanisms by which proteins find their correct target sites in living cells, and they play a very important role in elucidating the transcriptional circuitry that underlies cellular differentiation, oncogenesis and other human disease, and the response to environmental stress.

8:55-9:20 Regulation of Gene Networks and Cell Fate using Artificial Transcription Factors
Aseem Ansari, Ph.D., Assistant Professor, BioChemistry, University of Wisconsin - Madison

9:20-9:50 Deciphering the microRNA Code
Dr. Vera Atzorn, Marketing Manager, Actigenics 
In humans, protein expression largely depends on a “microRNA code” - a highly integrated functional network within which microRNA molecules work in a cooperative manner. MicroRNA show specific expression profiles in different normal / pathological cells and tissues, and are directly involved in numerous pathologies such as cancers, CNS diseases, metabolic diseases and certain viral infections. However, only a subset of these microRNA and their targets are known today. Actigenics has developed a powerful discovery platform that allows identification, validation and analysis of microRNA and their function in any cellular and tissue context. Well-designed algorithms allow identification of the set of functional targets for each microRNA and of the subset of microRNA acting in a coordinated fashion to regulate specific biological processes. Our technology and expertise offers innovative microRNA-based solutions aimed at understanding the function of microRNA in gene regulation in an integrated view of biological processes.

9:50-10:25 Coffee Break

10:25-10:55 Regulation of Gene Function by Man-Made, Vector-Based Intronic MicroRNAs (miRNAs)
Shao-Yao Ying, Ph.D., Professor, Cell and Neurobiology, University of Southern California
The microRNA (miRNA), a single-stranded regulatory RNA, suppresses expression of other genes based on its complementarity to the target genes. We have developed a man-made miRNA cassette, which is inserted in the intron region of a protein-coding gene and transcribed by tissue-specific RNA polymerase type II (Pol II) promoter. With computer-identified gene sequences and potential miRNA sequences, one can explore the functional significance of these sequences both in vitro and in vivo with targeted candidates, using the intronic miRNA cassettes. In addition, the significant role of a particular gene function in the signal pathway or a stage of development of diseases can be examined with this approach, particularly the downstream mechanism of gene regulation.

10:55-11:20 Decipher Microarray Experiments: From Pathways down to Gene Regulation Mechanisms
Martin Seifert, Ph.D., Vice President of Microarray Business & Collaborative Research, Genomatix Software GmbH
We present a new complete analysis strategy for microarray analysis, which starts with Affymetrix single probe signals and carries the analysis all the way through pathway analysis and the evaluation of gene regulation mechanisms on the molecular level. Single probe-based analysis avoids problems created by outdated probeset annotation. The complete analysis of a publicly available dataset: “Smoking-induced changes in airway transcriptome” (Gene Expression Omnibus) proved our approach to detect more significant genes while reducing the false positive rate by an order of magnitude. This became most evident in the subsequent pathway and gene regulatory analysis. As the concept is based on raw data files, already existing datasets can be re-analyzed allowing a substantial improvement of the ROI.

11:20-11:45 Clustering of Microarray Data via Incorporating Knowledge from Ontologies 
Ben Goertzel, Ph.D., Chief Scientific Officer, Research and Development, Biomind LLC
Hierarchical clustering is a very common approach to analyzing microarray data, popular in large part because of its simplicity and scalability. However, it has been found that the performance of hierarchical clustering may be significantly improved in some cases by altering the nature of its input. Instead of just giving it gene expression values as inputs, one may give it “amalgamated” expression values, corresponding to the average expression of all the genes known to fall into a certain functional category, based on existing knowledge resources such as the Gene Ontology. In many cases this results in clusters that better reflect the structure of the data, e.g., in studies of cancer data it has been found to better differentiate diseased tissue samples from controls. In general, the incorporation of ontological information into hierarchical clustering permits this common technique to provide more insightful information regarding gene regulation and other aspects of biological dynamics.

11:45-12:15 Revisiting an Affymetrix Reference Dataset: Retinoic Acid Induction of MPRO Cells
Sonja Steppan, Ph.D., Senior Research Scientist, Target Research, Bayer HealthCare
Microarray data for myeloid progenitor (MPRO) cell differentiation time series experiment was analyzed with a novel bioinformatics approach that combines single probe level detection with promotor inspection and pathway analysis. The number of significant expressed genes increased compared to MAS5 analysis, whereas the false positive discovery rate was decreased. Specially genes involved in cell differentiation were detected and coregulated genes were identified.

12:15-12:45 A Cross-Species Systems Biology Approach to Mammary Cancer: Identifying Functionally Conserved Pathways between Species
Jeffrey E. Green, Ph.D., Head, Transgenic Oncogenesis Group, Laboratory of Cell Regulation and Carcinogenesis, National Cancer Institute
Although many reports demonstrate the utility of using expression profiles for predicting the natural history of tumor development and outcome, there has often been a lack of overlap in the gene signatures identified between different studies. We have developed a novel method to compare datasets from tumors of different species and have used this to generate a cross-species predictor of estrogen receptor status in mouse and human breast tumors. Importantly, the addition of the mouse data to the human data significantly improved the class predictor as compared to human data alone when tested on an unrelated human dataset. Since more variance in gene expression exists due to the great genetic heterogeneity in the human population, the mouse data provided increased sensitivity in selecting genes that improved predictive power. The cross-species approach improves the identification of functionally important genes and networks which might otherwise be missed by current methods and can be applied to many different systems and analyses.

12:45-1:00 Q&A with the Morning Speakers

1:00-2:30 Lunch on Your Own or Luncheon Workshop (Sponsorship Available)

2:30-2:35 Chairperson’s Remarks
Thomas Werner, Ph.D.

2:35-3:05 Informatics-Based Vector Design to Minimize Inappropriate Host Interactions
Monika Wood, Senior Scientist, R&D Bioinformatics, Promega Corporation
While it is often assumed that reporter genes and vectors do not influence genetic regulation, this may not necessarily be valid. To address this problem, we have pioneered an informatics-based approach to designing synthetic reporter sequences that minimize potential influences on transcription-level regulation of gene expression. The resulting vectors have lower background transcription and show evidence of reducing expression artifacts.

3:05-3:35 GREF_GATA Model to Predict Androgen Receptor Binding Sites within the Human Genome
Albert Dobi, Ph.D., CPDR, Chief, Section for Gene Regulation and Bioinformatics; Senior Staff Scientist and Research Assistant; Professor of Surgery, Center for Prostate Disease Research, Department of Surgery, Uniformed Services University
Androgen hormone signaling is the main therapeutic target in prostate cancer. Therefore, the identification of genomic targets for androgen receptor is important for the development of better prognostic, diagnostic and therapeutic tools. Here we report a novel in silico model for the prediction androgen receptor binding sites. Our model is a trained composite model that is based on training datasets selected by biological function. This model demonstrates superior characteristics in predicting androgen receptor binding sites when compared to traditional weight matrix-based predictions.

3:35-4:00 Identification of Major Transcription Factors Responsible for Particular Cellular Regulatory Responses
Anton Yuryev, Ph.D., Director of Application Science Department, Ariadne Genomics
We describe the algorithms to identify major transcription factors responsible for the particular cellular regulatory response. The approach uses microarray expression data and the regulomics database of relationships between proteins. The most likely regulators are chosen based on the correlation of the differentially expressed downstream targets with the effect of a known corresponding regulation event in the regulomics database. Similar approaches can be used for metabolomics, transcriptomics and proteomic experiments.

4:00-4:20 Refreshment Break

4:20-4:50 Transcription Factor Blockade, Transcriptional Networks and Target Discovery: How a “General” Transcription Factor Mediates Specific Effects
W. Keith Jones, Ph.D., Assistant Professor, Pharmacology and Cell Biophysics, University of Cincinnati
Commonly, transcription factor activation is the result of upstream signaling cascade activity. It is the activation of particular transcription factors and co-factors in combination with the DNA binding site arrangement of promoters that results in the activation of specific downstream genes. When multiple genes are co-regulated by common factors, this constitutes a transcriptional network of gene expression. Such networks have specific effects upon normal physiology or pathophysiology that includes both direct effects of gene expression and indirect effects of signaling cascade feedback regulation. Specific transcription factor blockade can be used to discern the factor-dependent genes that are regulated in association with specific pathophysiological effects. The transcription factor NF-kappaB regulates a wide variety of biological effects in diverse cell types and organs, particularly stress and adaptive responses. Recently, it has become recognized that NF-kappaB regulates specific antithetical effects. In the heart, we have found that NF-kappaB is pro-cell death and mediates myocardial infarction after ischemia/reperfusion injury, and yet NF-kappaB is required for the cardioprotective phenomenon of late ischemic preconditioning and is pro-cell survival after chronic coronary ligation. To unravel the mechanism of how NF-kappaB mediates these paradoxical effects, we engineered transgenic mice with cardiac-specific blockade of NF-kappaB. We also describe novel and efficacious non-viral methodology for DNA decoy delivery and transcription factor blockade. These two complementary strategies allow us to determine the specific sets of NF-kappaB-dependent genes that are activated after different ischemic stimuli, in vitro and in vivo.

4:50-5:20 Generation and Use of Regulatory Gene Networks in Endothelial Cells
D. Steve Charnock-Jones, Ph.D., University Reader, Obstetrics and Gynecology, University of Cambridge and GNI Ltd.
The endothelial cells that line blood vessels play an important role in many serious conditions including cancer, rheumatoid arthritis and cardiovascular disease. The development of novel drugs that specifically regulate the key functions of these cells will lead to significant therapeutic benefits. To achieve this goal we have generated endothelial-specific gene regulatory networks using a large number of RNAi knockdowns and gene array analysis. We have used proprietary algorithms to infer the underlying regulatory network. This network reveals novel regulatory relationships and predicts likely control points for therapeutic intervention.

5:20-5:45 An Integrated Computational and Experimental Approach for Identifying A Compound Mode of Action
Guillaume Cottarel, Ph.D., President, Systems Biology, Cellicon Biotechnologies
A major challenge in drug discovery is distinguishing the molecular targets of a bioactive compound from the hundreds to thousands of gene products that respond indirectly to changes in the activity of the targets. Here, we present an integrated computational and experimental approach for identifying the gene products and pathways that are targets of a compound. This is achieved by filtering the mRNA expression profile of compound-exposed cells using a reverse-engineered model of the cell’s gene regulatory network. Using the method, we successfully predicted the molecular targets of multiple compounds, including a potential new anticancer compound, PTSB. PTSB inhibits growth in human small lung carcinoma cells and in the test organism (baker’s yeast). In follow-up experiments, we verified that PTSB acts on thioredoxin and thioredoxin reductase, the molecular targets predicted by our algorithm. These findings have validated the algorithm’s capabilities and facilitated investigation of a novel class of therapeutic compounds.

5:45-6:00 Chairperson’s Closing Comments