Preparing a human membrane and secreted protein-enriched cDNA library using PCR primers derived from a genomic database.

We describe here a strategy for preparing a human membrane and secreted protein (MSP)-enriched cDNA library based on human MSP- and non-MSP-encoding cDNA sequences in the databases. The signal peptide parts of the MSP-encoding cDNA sequences, which currently comprise about half of the estimated total number in humans, were analyzed for common patterns. These patterns form a 'minimal' set of polymerase chain reaction primer candidates of length varying from 9 to 21 nt. The products stemming from each primer candidate were determined and the results allowed us to obtain an 'optimal' mixed-length primer set. Ninety-six percent of the primers in this set were predicted to yield </=10% undesired products, and the desired MSP-cDNA products could be easily separated by gel electrophoresis. The present analysis establishes a methodology for preparing a cDNA library that enables the analysis of individual MSPs. This methodology may also help identify new MSPs. As many cell regulatory processes are mediated by secreted proteins and their membrane-bound receptors, the preparation of a MSP-enriched cDNA library should benefit research on MSPs.