Cloning in yeast: an appropriate scale for mammalian genomes

Abstract

The ability to clone fragments of DNA in prokaryotic systems has caused an explosion in our understanding of gene structure and flmction. This technology has been adopted as a major tool by branches of biology as disparate as taxonomy and neurobiology. It has spawned a multi-million dollar industry and generated the capacity to deal with major problems (AIDS springs to mind immediately) in a way and. with a speed which would have been impossible 15 years ago. But with current methods a gap between cytogenetic or genetic scales of analysis and the scale of molecular doning remains, especially when tb~ object of study is a human system. In current studies employing E. coil as a host organism, the size of the DNA fragment that can he cloned is limited by the biology of bacteriophage lambda. Closing this gap is one reason for wishing to be able to clone large fragments of DNA 1'2. There are many examples of human diseases for which linked markers exist. In physical terms these markers may be separated from the gene by meffdbases of DNA. Elaborate schemes have been devised to cover these distances by several 'jumping' steps ~'4 but this approach appears technically complex and fails to isolate the DNA between take-off and landing points. Chromosome mediated gene transfer s capable of cloning DNA fragments that are megabases in length but the complexity of the host genome is a disadvantage, as is the lack of control over the size of fragments transferred. A further major problem with this approach is that rearrangements occL~r with relatively high frequency ¢. The second reason for wanting a large scale cloning method is that some functional units are very large; for exm-aple, factor VIII covers 190 kbp or about 0. I% of the humm~ X chromosome 7, the Duchenne musmegahase or more s, and the distances involved in immunogiobulin supergene family reanangements have so far been too great to be determined. The third reason for cloning on this scale is to provide a starting point in the human genome sequencing project: an ordered set of clones seems logistically a simple place to begin. If fewer clones need to be ordered initially, fewer resources will be needed, the number of errors will monitor be lower and they will be easier to detect. Burke, Carle and Olsen 9 report an elegant approach to the cloning of large DNA fragments. By forsaking E. coli for Saccharomyces cerevisiae they have been able to exploit the availability of cloned telomeres m and centromeres ~1 to produce a 'yeast artificial chromosome' (YAC) that contains human DNA fragments hundreds of kilobases in size. A bacterial plasmid that carries two yeast telomeres, a yeast centromere and autonomously replicating sequence (ARS), with selectable markers and a selectable cloning site, is used to generate two 'arms' (Fig. 1). One arm contains a teiomere and a selectable marker and the other contains another selectable marker m~d both a centromere and a telomere. Ligation of large fragments of DNA to these arms, followed by transformatio~ into yeast and selection gives rise to