1. The study of tumor-suppressive mechanisms of transcription factor XX in pancreatic ductal adenocarcinoma show that XX induces the WNT signaling pathway and revealed a subset of 36 WNT activation targets that were down-regulated by XX. The Bioinformatics core helped Dr. Itkin-Ansari with mechanistic biological network reconstruction to uncover pathways from tamoxifen-induced XX to 30 of those targets (A). These are significantly enriched with pancreatic cancer genes and metastases-related genes, showing the role of XX as a potential tumor suppressor. Knowledge-driven pathway reconstruction helped elucidate the mechanisms employed by XX, similar to p53, p16, and SMAD4, or opposite of the oncogenic mechanism of K-RAS (B). Many XX pathways lead through p21, and subsequent inhibition of E2F1 and c-Myc, which leads to down-regulation of multiple cell cycle genes shown by microarray data (overlaid). Collectively, these results underline a tumor suppressive and potentially therapeutic role of XX induction.
2. We helped to research teams of Dr. A. Strongin and Dr. M. Perucho to focus on finding novel genetic and epigenetic markers of carcinogenesis in non-MSI colorectal cancers (CRC) by whole exome sequencing of cancer biopsies and matching normal tissues. IDM has accomplished extensive bioinformatics analysis of high-throughput sequencing data including genomic mapping, single nucleotide variation (SNV) detection and global statistical data analysis. These efforts led to discoveries of novel somatic mutations and SNVs associated with colorectal cancer. The validation of biological significance of novel CRC-specific mutations and SNVs is currently ongoing. This novel information will allow our researchers to dissect more efficiently pathological consequences of SNV in CRC.
2. One ongoing collaboration with Barbora Malecova and Lorenzo Puri has generated a statistical analysis of many chIP-seq data sets. This has shed light on the molecular mechanism of action of how the phosphorylation of p300 acetyltransferase and transcription coactivator modulates activity of a discrete set of promoters of DNA damage responsive genes in response to genotoxic agent. We have contributed in-depth analyses to studies featuring variant analysis and mutation mapping, RNA-seq and chIP-seq. We can align and QC NGS datasets, normalize data files and determine differentially expressed genes, call chIP-seq peaks using a variety of popular algorithms, and annotate results sets.
3. We contributed to the development of a new online stem cell analysis tool, called PluriTest. This was developed in part as a collaboration between The Scripps Research Institute and Bioinformatics Shared Resource at Sanford-Burnham. The tool is a quality control diagnostic test that will make it much easier and faster for researchers to determine whether their cell lines are normal pluripotent cells.PluriTest is available on website pluritest.org; you simply load the raw .idat file from an Illumina HT12v3 or HT12v4 array and the tool does the rest. Details and data can be found in this March 6th Nature Methods paper.
4. We performed data analysis to help establish the role of Siah2 in tumorigenesis and metastasis by HIF-dependent and -independent mechanisms (Dr. Z. Ronai, Signal Transduction). Critical computational support from IDM led to a novel paradigm for formation of neuroendocrine tumors, which involves the ubiquitin ligase Siah2 and the downstream effector HIF-1α in cooperation with FoxA2 that transactivates key proteins implicated in development of NE and NED. Study supported by the comparison/analysis of microarray data with cancer datasets publicly available in the NCBI GEO repository. IDM support included microarray data analysis (data generated in the Genomics Facility), clustering, cross platform analysis, pathway mapping, comparison of data to public data sets, preparation of illustrations and description of the results for the publication (Qi J, Nakayama K, Gaitonde S, Goydos JS, Krajewski S, Eroshkin A, Bar-Sagi D, Bowtell D, Ronai Z. The ubiquitin ligase Siah2 regulates tumorigenesis and metastasis by HIF-dependent and -independent pathways. Proc Natl Acad Sci U S A. 2008 Oct 28;105(43):16713-8. 2008; Qi J., Nakayama K., Cardiff RD., Borowsky AD., Williams R., Krajewski S., Carpenter P., Mercola D., Bowtell D. and Ronai Z. Siah2-dependent concerted activity of HIF and FoxA2 regulates formation of neuroendocrine prostate tumors. Submitted).
5. We assisted in the analysis of Mass Spectroscopy data (from the Proteomics core) to determine new protein markers for colon cancer (PI: Jeffrey Smith, Tumor Microenvironment). Results from this study show that inhibition of FAS, by either knockdown with small interfering RNA or inhibition with the small molecule drug orlistat, leads to activation of the receptor-mediated apoptotic cascade (caspase-8-mediated) and ultimately to cell death. However, knockdown of two enzymes upstream of FAS, acetyl-CoA carboxylase- and ATP-citrate lyase, fails to activate caspase-8 or to elicit apoptosis in tumor cells, even though palmitate synthesis was suppressed. Differential gene analysis revealed the unique apoptotic effect of FAS inhibition was due to up-regulation of DDIT4, a stress-response gene that negatively regulates the mTOR pathway. These findings indicate that the unique effect of inhibition of FAS results from negative regulation of the mTOR pathway via DDIT4. In this study IDM’s role was microarray data analysis, NMF data clustering, design of custom computer code to filter data by special criteria, pathway mapping, Gene Set enrichment analysis and preparation of figures and text (Knowles L., Yang C., Osterman A., Smith J. Inhibition of Fatty-acid Synthase Induces Caspase-8-mediated Tumor Cell Apoptosis by Up-regulating DDT4 J. Biol. Chem., Vol. 283, 46, 31378-84, 2008).
6. We provided help with the statistical analysis and clustering of gene expression data that helped in the identification of a mechanism that links ERK with JNK signaling in human melanoma (PI: Ze’ev Ronai, Signal Transduction). Constitutively active ERK increases c-Jun transcription and stability, which are mediated by CREB and GSK3, respectively. Subsequently, c-Jun increases transcription of target genes, including RACK1, an adaptor protein that enables PKC to phosphorylate and enhance JNK activity, enforcing a feed-forward mechanism of the JNK-Jun pathway. Activated c-Jun is also responsible for elevated cyclin D1 expression, which is frequently overexpressed in human melanoma. Our data reveal that, in human melanoma, the rewired ERK signaling pathway upregulates JNK and activates the c-Jun oncogene and its downstream targets, including RACK1 and cyclin D1. (Lopez-Bergami P, Huang C, Goydos JS, Yip D, Bar-Eli M, Herlyn M, Smalley KS, Mahale A, Eroshkin A, Aaronson S, Ronai Z. Rewired ERK-JNK signaling pathways in melanoma. Cancer Cell. 2007 May;11(5):447-60)
7. We provided a critical informatics support to the project of Dr. A. Strongin (Tumor Microenvironment Program) that led to the identification of 11 genes coordinately expressed with MT1-MMP protease in human tumors of 14 cancer types. Genome-wide expression profiling of MT1-MMP-overexpressing versus MT1-MMP-silenced cancer cells and a further data mining analysis of the preexisting expression database of 190 human tumors of 14 cancer types led us to identify 11 genes, the expression of which correlated firmly and universally with that of MT1-MMP (P < 0.00001). The results suggested that transcriptional reprogramming of the specific downstream genes, which themselves are associated with tumorigenesis, represents a distinctive "molecular signature" of the proteolytically active MT1-MMP. This study also identified a need for an improved cancer/normal gene expression tissue atlas. Investigation of the available resources revealed the Neurocrine tissue atlas and the ~1000 array GSK cancer cell line atlas as potentially valuable data sources. IDM efforts have made both of these large resources available to Burnham researches. The 1000 array GSK data set is now served by the in-house IDM-developed MESA data server (Rozanov DV, Savinov AY, Williams R, Liu K, Golubkov VS, Krajewski S, Strongin AY. Molecular signature of MT1-MMP: transactivation of the downstream universal gene network in cancer. Cancer Res 68(11):4086-96 (2008).
8. We contributed to the study of direct targets of the essential transcription factor Ets2 (PI: Bob Oshima, Tumor Development). The ets2 inactivation increased expression of a subset of differentiation-associated genes and decreased expression of several genes implicated in TS self-renewal. Thus Ets2 contributes to the regulation of multiple genes important for maintaining the undifferentiated state of cells and as candidate signals for embryonic development. In a related study, IDM performed a comparison of tumor expression data sets from different microarray platforms. It was demonstrated that the application of median rank statistics is a powerful tool for cross-platform comparisons. Combinations of several tools allowed IDM to generate gene lists consistent with other types of evidence (Wen F, Tynan JA, Cecena G, Williams R, Múnera J, Mavrothalassitis G, Oshima RG. Ets2 is required for trophoblast stem cell self-renewal. Dev Biol. 2007 Dec 1;312(1):284-99
6. We served as a key co-developer of the web-site, database and protein annotation pipeline for the Center on Proteolytic Pathways (PIs: A. Godzik, Apoptosis & Cell Death Research and J. Smith, Tumor Microenvironment). Proteases and their substrates play many important roles in cancer development. In this project IDM participated in development of The Proteolysis MAP (PMAP, http://www.proteolysis.org), a user-friendly website intended to aid the scientific community in reasoning about proteolytic networks and pathways. PMAP is comprised of five databases, linked together in one environment. The foundation databases, ProteaseDB and SubstrateDB, are driven by an automated annotation pipeline that generates dynamic ‘Molecule Pages’, rich in molecular information (see picture with example). The IDM role was to design, to implement (part) and to lead the development of the web site and databases. Examples of cancer-related proteolytic events can be found by link http://cutdb.burnham.org/relation/pmap2cutdb?disease=cancer (Igarashi Y., Heureux E., Doctor KS., Talwar P., Gramatikova S., Gramatikoff K., Zhang Y., Blinov M., Ibragimova SS., Boyd S., Ratnikov B., Cieplak P., Godzik A., Smith JW., Osterman AL., and Eroshkin AM. PMAP: databases for analyzing proteolytic events and pathways Nucl. Acids Res. 2009 37: D611-D618)
Information and Data Management Projects
For Genomics Shared Resource: analyzed microarray data and developed user-oriented MESA database and tools for handling functional genomic data: (http://bsrweb.burnham.org/metadot/index.pl).
For High Content Screening Core: In collaboration with San Diego Supercomputing Center we increased the computing and storage resources of that core. Using the national "TeraGrid" (teragrid.org) supercomputing resource this allows for raw data from high content screens – often 1 TB or more per screen – to be shared easily with other labs.
For Proteomics Shared Resource: Analyzed core-produced PI data; implemented LIMS system for managing, analyzing, and sharing large volumes of mass spectrometry data.
For Animal Facility: Added statistics and Improved functionality of the main database providing a better access to tracking animal inventory, breeding and maintenance (~30 labs have asked and received accounts). An improvement to Kaplan-Meier statistics for mouse survival.
A more detailed description of two selected projects is below.
Animal Facilities Informatics. The Burnham Animal Facility users had requested an improved informatics solution to keep track of breeding and maintenance of the ~7,000 cages of mice for 38 labs using mice at Burnham. The database in use is "Facility" a commercial product that is not highly user-friendly. IDM created a web-based interface that allowed lab-users to easily see statistics summaries and data critical to making important decisions, such as which mice to keep and how many breeders are required. The solution was implemented as web-interface that connects directly to the Facility database routinely used by animal technicians. This development yielded a system allowing researchers to access the same data via vastly improved interface, which is (i) focusing on lab users’ core needs, (ii) simplifying access via the web; (iii) integrating form-based requests; (iv) securing data by limiting access to the more open Facility database.
How did it work? (1) Wide acceptance, over 100 users, average of 4,000 queries a month; (2) time delay between AF work done on mice and reporting in the database has DECREASED – thus communication has improved; (3) lab PIs and researchers appreciate the service and rely on the information to make informed decisions. The list of Cancer Center Labs using the system includes 30 groups
Proteomics Facility Informatics. The proteomics shared resource required support to build a custom sample management and analysis solution to improve their customer support and make data management more efficient. They employed the IDM to perform a comparative expert analysis of potential commercial and public domain software tools. A choice of LabKey Server from the LabKey Corporation made based on this analysis provides the Proteomics facility with all needed capabilities including web-based data management, data sharing, on-line reports and discussions (illustration of the LabKey interface is shown on Figure.). In the near future IDM will assist in deploying this system for the entire Cancer Center and expanding its functionality towards management of microarray data.