Julia E. Parrish. Evan E. Eichler, Beth A. Firulli, A. Craig Chinault, Mark Graves, Andrew Arenson, Cheng Chi Lee, and David L Nelson
Department of Molecular and Human Genetics Baylor College of Medicine, Houston, TX 77030.
The distal long arm of the human X chromosome is one of the most gene-rich regions of the genome; however, of the 22 genetic disorders with Xq28 linkage, 14 remain uncloned. The majority of Xq28 has been assembled into YAC contigs (Palmieri et al., 1994, Genomics 24: 149158). In order to improve the level of resolution of the physical and transcription maps in this region, we have implemented a two-step strategy utilizing the Lawrence Livermore flow-sorted X cosmid library. YACs are selected to provide 1-3 fold coverage of the region. The YACs are subjected to long-range Alu PCR using six combinations of primers, yielding a range of products up to 13 kb. Products from each PCR reaction are labeled separately, then pooled and hybridized to the cosmid library. Data are entered directly into an in-house database using a graphical interface, which reduces both error and time spent decoding the coordinates of positive signals. Data are then transferred into a relational database, which allows comparison of the results with other screening data generated at Baylor. Cosmids are placed into "bins" defined by hybridization to one or more YACs in the contig. To date, 845 cosmids have been placed into 28 bins in a region of about 4.5 Mb which extends from the FRAXE region to BGN, and 125 cosmids have been placed into 5 bins within the ~2 Mb interval between Factor VIII and the telomere. These efforts provide a substantial degree of coverage of Xq28 in cosmids.
The second step of our approach involves comparison of our data with that generated by Cheng Chi Lee's cDNA/cosmid reciprocal probing strategy. Briefly, his approach is to use pools of individually arrayed cDNA clones (from heart and placental libraries) to probe the cosmid library. Positive cosmids are then used to pursue individual cDNA clones. The pooled cDNA to individual cosmid data are present in the in-house database. Multiple cosmids within a bin which are found to be positive for a given cDNA pool provide evidence for a gene within that bin; moreover, these data allow the cosmids which likely contain that gene to be given first priority in further analysis. We have pursued five such correlations to the single cDNA clone level; additional correlations indicate the presence of two genes, expressed in heart, which lie within the critical region for Barth syndrome.
Supported by grants DE-FG05-92ER614()1 and DE-FG03-94ER61830 from the U.S. Department of Energy and a Center grant from the NCHGR of the NIH (NIH 5P30 HG00210) to DLN.