Fluctuation tests of the Luria-Delbruck type have been used to examine the frequency of spontaneous variations in cultures of diploid, tetraploid, and octaploid Chinese hamster cells. Resistance to 8-azaguanine and al- tered response to thermal shock were chosen as marker systems. Mutation frequencies of approximately 10-e and have been estimated for heat resistance and azaguanine resistance respectively. Rates of mutation for these markers remain constant or decline slightly in cells with increasing numbers of chromosome sets. This trend is not in accordance with expectations based on the assumption of dominant, co-dominant, or recessive changes at gene or chromosome levels. It is suggested that at least some variations may arise in somatic cells by stable shifts in phenotypic expression rather than by changes in genetic information.
Cytogenetic studies on cell cultures have focussed on variations of two principal types. One class consists of mutant characters first recognized through hereditary defects in the intact organism, and which later were found to be expressed in vitro as well. Galactosemia, acatalasia, and orotic aciduria are typical examples of this group, which provides an array of stable markers for study in culture systems (see Krooth and Sell, '70). The second category of lariations has a less clearcut genetic backg.-ound, and is made up of stable changes that arise de novo in isolated cell populations. Such variations appear within clonal cultures, and are mitotically transmissable in the absence of selective agents. Drug resistance is one well-known erample; changes in nutritional requirements (Puck, '70) or in thermal sensitivity (Harris, '67) are other illustrations.
Variations that arise de novo in somatic cell cultures have usually been interpreted as the result of mutation at gene or chromosome levels. The evidence which underlies this assumption has accumulated primarily through fluctuation tests, patterned after the classic experiments of Luria and Delbruck ('43) in microbial systems. In such assays the variance for any given marker in replicate samples from a single culture is compared with that of samples taken from a series of independently maintained sublines. If in the latter the observed