MYRIAD GENETICS’ YEAST TWO-HYBRID SYSTEM BACKGROUND INFORMATION

Background on Myriad's Y2H System: Myriad's approach involves the development of novel reagents that optimizes the co-expression of both the DNA binding domain and activation domain hybrid proteins in a single yeast cell. Additionally, the application of Myriad's advanced robotics and bioinformatics capabilities further developed the technology into a robust and high fidelity platform that has been termed ProNet (for Protein Network). The success of the ProNet strategy relies upon using a large number of overlapping DNA-binding domain fusions for each starting protein. DNA-binding domain constructs are prepared in 96-well format using liquid handling robotics, and their quality is ensured by DNA sequence analysis. Verified constructs are screened against as many as three different activation domain libraries. Libraries have been generated from a variety of different human and mouse tissue and cell types. Each library is generated from random-primed, directionally cloned cDNA and is typically composed of over 5 million independent clones with an average fragment size of 600 to 900 base pairs. Thus, every ProNet experiment screens a bait protein fragment against multiple overlapping regions of each library protein. Existing libraries from human tissue or cell types include: adipose; breast cancer cell lines, brain, endothelial cells, heart, hippocampus, hypothalamus, hypoxic cells (acute and chronic), kidney, liver, macrophage, placenta, prostate cancer cells, skeletal muscle, small intestine, spleen, and thymus. Libraries generated from mouse tissues include: mouse B cells, cardiac myocytes, and total mouse embryo.

In addition to the libraries listed above, custom activation domain libraries can be generated by request. Importantly, problems with false positive results from the Y2H system have been handled in the ProNet process. The use of multiple reporter genes and a panel of specificity tests reduce the level of false positive results substantially. Furthermore, their analysis of over 50,000 Y2H screens has identified the proteins that behave as false positives and permits the elimination of any that remain in the data sets.

The ProNet process has been engineered to achieve high quality and throughput. Specialized informatics tools manage the process in the laboratory and allow the end-users to review, understand and use the data. End users (biologists) can access the ProNet data using a web-based interface. This interface is organized into specific home pages for each protein and interaction.

Upon the initiation of a research collaboration, target proteins are submitted by the collaborator for analysis. The project manager designs and orders DNA-binding domain constructs for each target protein. Access to the Myriad web-based interface (customized for each project) is initiated at the outset so that the collaborator may follow the progress of the project and also view interaction data curated from the literature. As data are generated, the project manager will provide a description of the interactions, incorporating relevant information from the scientific literature. These descriptions are posted on the web site weekly intervals. Quarterly meetings and frequent phone conferences are opportunities to guide the research and to exchange additional project information.