Diabetes Genome Anatomy Project

Core Labs > UMASS Genomics Core

UMASS Genomics Core

Core Director: Michael Czech, Ph.D. (U. Mass. Medical School)

Specific Aims

The UMass Genomics Core will provide microarray technology to support DGAP activities in gene discovery and expression profiling related to mouse genes in fat and muscle cells. These tissues uniquely express the GLUT4 glucose transporter and respond to insulin with translocation of intracellular GLUT4 to the plasma membrane. Preliminary experiments using suppressive subtraction hybridization have generated a database of genes that exhibit high expression levels in both fat and muscle, and low expression in fibroblasts. This database contains both known and uncharacterized genes potentially involved in insulin signaling. Data obtained by microarray analysis of primary and cultured fat cell and muscle gene expression in this Core will greatly expand our efforts at discovering and characterizing such genes. Since rapid functional analysis of discovered genes is crucial to fully exploiting genomics and proteomics technology, a high priority of the Core is to develop new technology for selective gene silencing in insulin-responsive cells. The major goals of this Core are as follows:

Aim 1.
Provide analyses of Affymetrix GeneChips to UMass DGAP projects, as well as expertise in data mining, database annotation, and dissemination of DGAP databases to the worldwide research community. Expertise for preparation of cRNA probes and GeneChip analysis is already established in this UMass Core facility, and efforts are underway to coordinate its activities with the Harvard DGAP Genomics and Bioinformatics Cores. In this regard, experiments will be conducted to test the hypothesis that similar gene expression profiles are obtained from samples prepared and analyzed at the two sites.

Aim 2.
Develop methods using siRNA oligonucleotides to silence selected genes in cultured 3T3-L1 adipocytes to test their potential role in GLUT4 translocation in response to insulin and other fat cell processes. The Core will extend recent exciting preliminary data to develop and optimize experimental conditions which promote ablation of gene expression through application of specific siRNA oligonucleotides to cultured adipocytes. This will be coupled to the use of a semi-quantitative method to estimate movement of exofacial-tagged GLUT4 to the plasma membrane in such cells.

Initial Studies and Results

A primary function of the UMass Genomics core will be to analyze expression profiles of cRNA samples produced from experiments using murine model systems that differ in insulin-responsiveness. Currently, the most complete mouse GeneChip available is the Affymetrix mouse U74 series, consisting of 36,000 genes and EST sequences on three separate GeneChips. The U74A chip contains sequences (~6,000) in the Mouse UniGene database (Build 74) as well as ~6,000 EST clusters. The U74B and U74C GeneChips each contain ~12,000 EST clusters. As an initial approach to identify genes potentially involved in insulin signaling, we recently screened the Affymetrix Mouse U74 GeneChip series with cRNA probes made from mouse muscle and both 3T3-L1 adipocytes and 3T3-L1 fibroblasts. Figure 1 shows typical results of these hybridizations for the adipocytes vs fibroblasts for each U74 GeneChip. The large number of data points that reside above and below the diagonals indicates that there are many genes whose expression is either up- or down-regulated in response to adipocyte differentiation. Genes that were found to be highly expressed in adipocytes on the U74 chips include genes such as Glut4, AdipoQ, and adipsin, which have been previously identified to be highly expressed in 3T3-L1 adipocytes upon differentiation. Furthermore, signals of the majority of adipocyte-specific genes on the U74C EST chip are significantly lower in expression than most of the adipocyte-specific genes on the other two chips. These adipocyte-specific genes on the U74Cchip might represent very low abundance gene transcripts that are up-regulated upon adipocyte differentiation but escaped previous identification.

U74A	U74B	U74C

Figure 1. Gene Expression Profies of 3T3-L1 fibroblasts and adipocytes using the Affymetrix Mouse U74 series GeneChips. The three mouse U74 genechips were probed with biotin-labeled cRNA probed from mRNA isolated from 3T3-L1 fibroblast and adipocytes. Each chip is composed of about 12,000 different probe sets. The log scale scatter plots show the average difference calculation of signals that hybridize to each probe set. The Y axis displays the relative expression of genes and ESTs in 3T3-L1 adipocytes, while the X axis displays their relative expression in 3T3-L1 fibroblasts.

Genes expressed selectively in fat and muscle and those exhibiting expression patterns that respond to conditions that change insulin sensitivity become candidates for further functional analysis to determine whether they have a role in insulin signaling and diabetes. This has been a major obstacle to exploiting genomics methods since many more differentially expressed genes are often identified than can be reasonably studied one by one. Recent work, including groundbreaking studies by UMass investigators Craig Mello and Phillip Zamore, has shown that gene silencing by interference RNA can be an effective way to quickly screen the requirement of a specific gene for a biological function. The discovery that short sequences of double stranded RNA can act to silence specific genes in cultured mammalian cells has opened the way for potentially testing with high through-put whether any gene within a large set of genes is required for a particular function. Our initial efforts to translate this technology to insulin-sensitive 3T3-L1 adipocytes completely failed, using exactly the conditions that cause genes to be "knocked out" virtually completely in 293 and other cells. However, by methodically varying conditions and entry modes we have recently been able to successfully and specifically inhibit expression of a test gene in cultured adipocytes, which provides high confidence we can further develop and improve this method.

Preliminary results on siRNA-mediated gene silencing in 3T3-L1adipcytes can be found in our recent publication, Bose et al. (2002) Nature 420: 821-824. Further work on this method, which we have submitted for publication, shows even more highly effective gene silencing of Akt1 and Akt2 in cultured adipocytes.

Related Projects

• Project 3: Anatomy of Gene Expression in Insulin Resistant States