The amplification of oncogenes is an important determinant of tumor progression in humans. We have obtained substantial evidence by fluorescence in situ hybridization (FISH) that amplification of the model dihydrofolate reductase (DHFR) gene in both CHO and human cell lines is initiated by chromosome breaks resulting from sister chromatid fusions. Major unknowns include the underlying causes of the breaks and whether they relate to telomere maintenance, the mechanism of chromatid fusion, and the processes that trim and homogenize repeating units during amplification to higher copy number. Any of these steps is a potential target for chemotherapy, and our long-range goals are to understand the molecular mechanisms operating at each step.
Specific aims of this proposal are: l) to define the mechanism that fuses chromatids after an initial chromosome break by isolating and characterizing the fusion product(s) of well-defined broken chromosome ends in first-step amplificants; an efficient homologous recombination approach will be used to introduce a cassette of rare-cutting restriction sites just downstream from the DHFR gene, and breaks will be elicited by introduction of the relevant restriction enzyme in trans; 2) to determine whether transient unmasking of telomeres can lead to gene amplification; a dominant-negative telomere-binding TRF2 protein will be over- expressed in cells with a DHFR-proximal telomere to induce end-to-end chromosome fusions; the frequency of amplification will then be determined, and rearrangements will be analyzed by fluorescence in situ hybridization (FISH) and restriction mapping; 3) to determine the extent to which break-induced replication can occur in CHO cells, as a possible model for homogenization and trimming of amplicons; we will determine whether the structures of truncated DHFR genes that have been restored to wild-type while retaining the original deletion junction have done so by an extensive gene conversion event; and 4) to test the hypothesis that break-induced replication can initiate amplification or can shorten, homogenize, and amplify initially large, heterogenous amplicons; FISH analysis will be used to divide the cell lines isolated in Aim 2 into those that obviously underwent initiating bridge-breakage-fusion cycles from those that appear' to have amplified the DHFR gene in loco; restriction mapping will reveal whether those amplifying in loco did so by rolling circle replication in the absence of an initiating bridge-breakage-fusion cycle.