Diabetes Genome Anatomy Project

Home > Core Labs

Core Labs

Samples from all projects will be subjected to a systematic, standardized and coordinated analysis in the four core laboratories. Gene expression will be assessed using identical protocols in two distinct core labs (one at Joslin and the other at the University of Massachusetts) using Affymetrix oligonucleotide arrays. Selected samples will be assayed in both laboratories to assess reproducibility and develop standardization for gene profiling. In our preliminary studies, we have already begun such a comparison, which illustrates the critical need for standardization and cross referencing if one is to maximize the value of the information obtained in this and other studies and link them in common databases (see discussion in Informatics Core and Joslin Genomics Core). Although various methods can be used to detect and quantitate mRNA levels, we have chosen to focus on the use of Affymetrix high-density oligonucleotide array technology, since this technology is increasingly widely available, reasonably sensitive, allows for simultaneous, quantitative analysis of expression of tens of thousands of genes in both rodents and humans, and is gradually becoming the de facto standard for use in laboratories throughout the world. Thus, one goal of the DGAP is to provide protocols and standard samples to allow for standardization across all labs using this approach for diabetes-related studies. Over the past year, both core laboratories have gained experience with the Affymetrix oligonucleotide arrays for expression screening. We find that the technique routinely detects expression of a gene present at 1:100,000 or less, that the system is reproducible, and that using replicates, it is possible to detect significant changes as small as 1.2-fold. We have also had the opportunity to compare this approach to others and find it very robust. From 50-100 mg muscle samples, sufficient RNA is obtained to do at least two Affymetrix analyses using the U74 (mouse) or U95/U133 (human) chip series. Using only the A chip, over 2000 genes/ESTs are detectable, and as noted in Project 1, several hundred of these exhibit changes in gene expression. One limitation of this technique is determining how best to analyze the myriad of results and determine statistical and biological significance. The Informatics Core has extensive experience in this area and will apply multiple techniques as described in Section D. This will also allow for categorizing genes based on extent of regulation, consistency of change, and correlation with metabolic or genetic status. For selected genes, the differences in expression will be confirmed using quantitative real time PCR (PE-ABI 7700) or by northern blot analysis of the same tissue samples. In addition, the Joslin Gene Expression Core will ultimately help develop "diabetes-specific DNA chips" that not only include insulin and diabetes regulated genes, but also their splice and sequence variants, that can be applied to the study of large populations groups, as well as animal and cellular model systems. For some cDNAs and ESTs which show significant changes in expression, but for whom the role of the gene product in insulin action or insulin resistance is unclear, we will use RNA interference to reduce expression and assess the biology of the gene product.

The sequencing will be performed in cooperation with the Center for Genome Research at the Whitehead Institute. Sequencing will focus on those genes or ESTs defined as good candidate genes by being highly regulated in response to insulin or diabetes. The Center for Genome Research will provide up to 10 kb of genomic sequence on 100 genes per year (for smaller genes this will comprise the whole gene, for larger genes, sequencing will focus on the exons, splice junction and the promoter region. The goal will be to determine the range of sequence and expression variation in a group of candidate diabetes genes, i.e. those highly regulated by insulin and/or diabetes. One proposed scheme for prioritization of this sequencing is illustrated here:

The Proteomics Core will provide a unique analysis of subcellular distribution and protein-protein interactions, as well as convention separation of protein components, and identification of separated proteins by mass spectrometry for adipose tissue from the human, rodent and cellular models for correlation with the gene expression data.

The Informatics Core will create and analyze the gene expression database; provide input into experimental design, standardization methodologies; and a format to interface the genetic, genomic and proteomic data with that of other NIH-funded efforts. The Informatics Core will also provide the information accrued in these studies in a timely fashion to all investigators in the field in a robust and highly annotated database. It is our belief that this will create important information and reagents that will be useful to all investigators in the fields of insulin action, insulin resistance and type 2 diabetes and will form a bridge with other genetics and genomics efforts to integrate this knowledge in the most effective way.