A visual intracellular classification strategy for uncharacterized human proteins

The human cDNA and genomic sequencing projects will result in the identific ation and isolation of some 140,000 genes, the majority of which lack: pred icted functions and for which the cellular localizations are not known. The identification and characterization of protein components of specific cell structures and machineries are essential steps not only toward defining fu nctions of genes but also toward understanding cell function and regulation . We describe here a new approach, termed PROLOG, which uses full-length cD NAs for systematic classification of novel proteins as a functional pointer . We have PCR-amplified 25 uncharacterized human genes and expressed the en coded proteins as GFP fusions in a human cell line. This pilot project has identified novel proteins associated with the nucleolus, mitochondria, the ER, the ER-Golgi-intermediate compartment (ERGIC), the GC, the plasma membr ane, and cytoplasmic foci. This visual classification approach may be scale d up to handle a large number of novel genes and permit the generation of a global cellular protein localization map. Such information should be valua ble for many aspects of functional genomics and cell biology. (C) 2000 acad emic Press Key