LiveWorld Transcripts

Bio Online presents

Panel of Scientists
Deciphering the Human Proteome

March 09, 2001

Scientists discuss the merits of a coordinated effort to decipher the human proteome and the development of technologies required to fulfill this challenge.

Page 1 of 7

Bio Online: Welcome to Bio Online's live discussion on tackling the human proteome. With the human genetic blueprint nearing completion, researchers are beginning to probe protein function and interactions on a genome-wide scale. But are the tools currently available adequate to tackle such a mammoth project? We have invited leading scientists to discuss the merits of a coordinated effort to the human proteome and the technologies required to fulfill this challenge. Now, I would like to introduce our panelists. We are very fortunate to have with us - Stan Fields, Ph.D. Professor of Genetics and Medicine, Howard Hughes Medical Institute, University of Washington. Eric Phizicky, Ph.D. Associate Professor of Biochemistry and Biophysics, University of Rochester. Jim Garrels, Ph.D. President and CEO, Proteome, Inc. Reudi Aebersold, Ph.D. Co-founder and Faculty, Institute for Systems Biology. I would like to start by asking each of you to describe your research and interest in proteomics.

Stan Fields: My interests have focused on developing technologies to analyze the function of proteins. In particular, my laboratory has exploited features of the yeast Saccharomyces cerevisiae to establish simple genetic assays. The most widely applied of these assays, the two-hybrid system, uses the modular nature of transcription factors to detect interactions between proteins. A related three-hybrid system, developed in collaboration with Marvin Wickens at the University of Wisconsin, detects RNA-protein interactions. Recently, we have extended these strategies to design functional screens that encompass an entire genome's worth of proteins. We developed a protein array format in which each predicted protein of an organism is expressed in a form that allows it to be easily assayed. The array format has been used to detect many interactions among S. cerevisiae proteins. It has also been adapted, in collaboration with Eric Phizicky and Elizabeth Grayhack at the University of Rochester, to a biochemical genomics approach that rapidly identifies proteins (and their genes) associated with biochemical activities. Additionally, we have used yeast to generate biosensors of ligand binding.

Eric Phizicky: The completion of the first draft of the human genome sequence gives us the potential to understand the precise protein content of every cell in a human being. It is also suddenly within our sights to decipher the function of every protein in the cell, and a number of methods have emerged to meet this challenge. Our lab has focused on a genomic approach to biochemical analysis. To this end, a genomic library of yeast strains was prepared, in which each strain expresses a different yeast protein attached to an affinity purification tag. With the purified genomic array of proteins, one can rapidly link any biochemical activity to its cognate gene. This array has been used to identify several different catalytic and binding activities, and can be used to find targets of drugs. A similar array could be developed to analyze the proteins of the human genome.

Jim Garrels: Many new technologies are starting to make possible the large-scale deciphering of the protein content of human cells. Our company, Proteome Inc, which is now a division of Incyte Genomics, has been building proteomic databases for the past 6 years. We started with several of the first sequenced organisms, yeast and worms, and systematically gathered up the world's knowledge of heir proteins into a new format - the BioKnowledge Library. Now we are doing the same thing for human, mouse, and rat proteins. To build the BioKnowledge Library, we have tapped into the largest source of proteomic data of all - the research literature, and our expert curators have recast the knowledge from huge numbers of scattered research articles into a new Internet-accessible format. Now proteomic researchers can interpret the results of each new experiment in the light of prior knowledge, even if the results involve hundreds of proteins each week. As the human proteome comes into full view, from further analysis of the human genome and through direct proteomic experiments, we will keep the BioKnowledge Library updated as the source for the latest and most comprehensive knowledge for all human proteins.

Reudi Aebersold: The study of the structure, function and control of biological processes is a central theme in biological and medical research. With the availability of the genomic sequences for a number of species a new research approach has been emerging that is based on the systematic and quantitative analysis of the genes expressed by a specific cell in a specific state. Our lab has focused on the development and initial applications of a technology to identify and determine the relative abundance of the proteins contained in a cell or tissue. This quantitative protein profiling technique is based on the use of a new class of reagents we call isotope coded affinity tags (ICAT), automated tandem mass spectrometry and sequence database searching. This method has enabled us to identify and determine the relative abundance of proteins extracted from cells representing different metabolic or differentiation states. Further current work in our lab is aimed at automating the analytical process based on the ICAT technology and to develop variations of the method that also allow us to quantitatively profile post translational modifications on a proteome-wide scale.

Page 1 of 7