) Genomic alterations underlie numerous human diseases, including inherited defects and cancer. Identification of disease genes is a multi-step, labor intensive process and is a rate limiting step in gene discovery. An approach to rapidly link gene sequence in a candidate disease interval with expressed proteins in the target tissue would facilitate identification of disease-associated genes. We propose a novel approach (""""""""epitope trapping"""""""") that utilizes the power of diversity and selection from combinatorial libraries to rapidly link DNA sequence in defined regions of the genome with expressed proteins in a target tissue. In this approach, an epitope library, constructed using sequence from the genomic region of interest (i.e. gene fragments), is screened against an antibody phage display library specific for proteins expressed in the target tissue. """"""""Trapped"""""""" epitope phages contain a DNA sequence corresponding to a coding (exon) region. The trapped epitope phages are selected, DNA sequenced and mapped back to the genomic region. A major windfall of this technique is rapid selection of epitopespecific antibodies that will be powerful reagents for functional analyses. We will establish proof-of-principle of epitope trapping for gene discovery using 5q3l sequence and an antibodyphage display library for proteins expressed in hemopoietic cells. Specifically, we plan to: 1) generate genomic epitope and cell-specific antibody-phage display libraries, 2) evaluate whether epitope trapping identifies known genes in a fully sequenced P1 from 5q3l, 3) determine whether epitope trapping efficiently identifies genes on large genomic intervals. We anticipate that epitope trapping coupled with antibody reagents will allow for high throughput identification of genes within a genomic region.
