April 14, 2006 | In a landmark for functional proteomics, researchers at Harvard Medical School have unveiled a publicly available library of proteins that are active in breast cancer tissue. The resource could help speed drug discovery efforts against the disease.
The researchers believe this is the first publicly available library of reliably characterized and expressible proteins for any human disease. The library consists of 1,300 complementary DNAs (cDNAs) corresponding to breast-cancer-related proteins. These proteins can be expressed in a model system that mimics human breast tissue, and as reported in February in the online Journal of Proteome Research, has already revealed some novel and surprising activities for several proteins.
Joan Brugge, chair of the Harvard Medical School Department of Cell Biology, and co-principal investigator of the Breast Cancer 1000 (BC1000) program, notes that breast cancer research has been stymied by the difficulty in ascertaining whether protein changes in cancer cells are cause or effect. “The systematic approach that we’ve enabled and demonstrated will allow researchers to track cancer-causing proteins in simulated environments, with the goal of learning how to impede them,” says Brugge.
“The availability of this collection will enable pilot experimentation and accelerate the development of faster techniques for studying breast cancer in a mammalian setting,” says Joshua LaBaer, director of the Harvard Institute of Proteomics, and co-principal investigator.
There was a dual approach to selecting the cDNAs for the BC1000 library. The first 200 genes were selected by breast cancer experts, along with another 50 genes known to be over-expressed in breast cancer. The remainder was identified using MedGene, a text-mining program developed by the Harvard Institute of Proteomics that searches content in the Medline database to pick out genes co-cited with a particular disease. The program ranks the significance of those gene-disease relationships based on citation frequency.
Initial screens on a subset of 265 cDNAs, introducing them into a breast cancer cell line and examining their effects on cell migration, proliferation, and morphogenesis, identified a number of cDNAs with putative roles in tumorigenesis, as well as migration-inducing activity in several surprising cDNAs.
The researchers have made the BC1000 library publicly available (see www.hip.harvard.edu/) and hope to develop and maintain a database for the completed clones as well as perhaps the experimental data.
Witt, A.E. et al. “Functional proteomics approach to investigate the biological activities of cDNAs implicated in breast cancer.” J Proteome Res 5, 599-610; 2006.