ABSTRACT
Normal mammalian cells rarely experience large-scale genetic rearrangements such as gene amplification, deletion, and translocation. There is general agreement that the earliest products in the amplification process are genetically unstable. A decade ago, Kaufman and Schimke reported that the products generated after a single step of methotrexate selection gave rise to heterogeneous dihydrofolate reductase (DHFR) gene expression and copy number, but the techniques employed did not enable a molecular or cytogenetic definition of the structures involved. The diversity of amplicon sizes and structures that can be generated from a single locus or from different loci are remarkable and have elicited proposals of numerous amplification mechanisms. The frequent localization of amplified DHFR genes to derivatives of the broken target chromosome may also derive from the existence of atelomeric chromosome ends.
