Normal diploid human and rodent cells lack a detectable frequency of gene amplification

Gene amplification is frequently observed in tumors and transformed cell lines. This phenomenon is known to contribute to the generation of drug-resistant tumor cells and quantitation of the event is believed to have prognostic value in several types of neoplasias. To date, most studies of gene amplification have used immortalized cell lines and biopsied tumor samples. In this study I examine the amplification potentials of primary diploid cells, both human and rodent, and quantitatively compare them to the amplification potentials of their transformed counterparts. I have used a strictly defined protocol (i.e., selection at a stringency of 9 X LD50) to measure amplification potential at two loci, the gene for the multifunctional protein containing activities for carbamoyl phosphate synthase, aspartate transcarbamylase, and dihydroorotase (CAD) and the gene for dihydrofolate reductase (DHFR). I find that the frequency of amplification in transformed cells is at least four orders of magnitude greater than that in normal cells. Out of 29 cell populations studied, the 7 diploid normal cell populations exhibited no detectable amplification frequency (limit of detection at 10(-8) whereas the 22 transformed cell lines demonstrated amplification frequencies between 10(-3) and 10(-7). These results demonstrate that a dramatic difference exists between primary diploid cell populations and immortalized cell populations in their ability to amplify genomic sequences and suggests a significant difference in genetic stability between these two cell types.