## Clone Parameters

3.5.1. Mean mwh Clone Size Class

Clones can be classified into size classes (i), delimited by powers 2i-1, according to the number of mwh cells they contain. For continuous exposure in the ideal case, the mean clone size class is i = 2 and the mean clone size is 2'-1 = 2 cells (geometric mean). In practice, the clones may be smaller or larger than theoretically expected. For compounds that are applied chronically, but are otherwise either unstable, bioactivated with delay, or inactivated rapidly during the last one to two rounds of cell division in the pupa, a correction of the estimated clone induction frequency may be appropriate according to the mean size of the clones (55).

Considering mwh clones from mwh single spots and from twin spots, it is possible to calculate the mean mwh clone size class (i). This figure represents the clone size class in which the majority of clones induced by a specific treatment is located.

To illustrate how it is possible to find out the mean mwh clone size class, we use some data extracted from Cunha et al. (19) as shown in Table 1. To calculate the mean clone size (geometric mean), one has just to apply the i, found in a specific treatment, in the formula 2-1. For example, the geometric mean of the clones found in the treatment with 0.05 mM of camptothecin is

3.5.2. Clone Induction Frequencies Per Cell and Per Cell Division

At the end of wing development, the wing consists of approx 30,000 cells. Wing development starts with some 30 cells in the embryo. We can estimate that there may be approx 10 rounds of cell division until metamorphosis. At each cell division round, the number of cells is doubled. Summing up the individual cell divisions gives C = 30 + 60 + 120 + 240 + 480 + 960 + 1920 + 3840 + 7680 + 15,360 ~ 30,000, which is also the final number of cells present in the adult wing. In other words, there are as many cells in the adult wing as there are cell divisions of precursor cells during development. There is one cell generation in passing from 30 to 60 cells, one passing from 60 to 120, and so on. In the last (i.e., the 10th generation), the primordium passes from approx 15,000 to the final 30,000 cells (Frei, personal communication).

The mwh clone frequency per fly makes it possible to estimate the induction frequency per cell and per cell division. An appropriate estimation of the induction frequency is obtained if the mwh clones per fly frequency is divided by the number of cells (48,800) present in both wings (see Table 2). We use 48,800 (24,400 per wing) instead of 60,000 cells (2 x 30,000â€”considering both wings), because in screening for wing spots, we do not examine all the cells in a wing; there are approx 24,400 cells in the wing area we inspect for spots (see Fig. 1).

It seems desirable to estimate clone frequencies in the SMART as induction frequencies per cell and per cell generation. It has been proposed that, depending on the time of induction, such frequency determinations should include a clone size correction. However, if the interpretation is correct, that small clone size in balancer heterozygotes reflects the presence of a chromosomal deficiency and not, or not only, a late time-point of induction in the course of development, a clone size-dependent correction of the clone induction frequency may be meaningless or may even falsify the result. Such would not only be the case for balancer heterozygotes, but to a certain extent also for inversion-free individuals, because in the latter, the same chromosomally aberrant clones are to be expected in addition to those produced by recombination. Cautious use of the clone size correction is therefore suggested, because in the case of particularly small clones, clone size-corrected induction frequencies may be under-estimations, mainly in balancer heterozygous flies. In the case of particularly large clones, however, the uncorrected frequencies may be underestimations. In critical cases, therefore, one would probably determine both values to indicate the possible range of these estimates, as indicated in Table 2 (55).

### 3.5.3. Quantification of Recombinagenicity

The relative frequency of twin spots may give some idea of the recombin-agenicity of a compound. Genotoxic chemicals can give quite different results in this respect (33). Under the assumption that mitotic crossing over is proportional to the physical distance on the chromosome between the centromere and marker genes, one would expect approx 50% of twin spots (recombination between flr3 and the centromere) and 50% of mwh single spots to be caused by mitotic crossing over (9).

Some data have shown that twin spot identification depends on clone size. Ramel and Magnusson (33) have already pointed out that for chemicals producing predominantly small spots, the lack of the flr3 genotype expression in small clones leads to considerable biases, because small twin spots cannot be readily identified.

For an unbiased evaluation of recombinagenicity, it is therefore preferable to compare the mwh clone frequencies in the two genotypes mwh/flr3 and mwh/ TM3 (see Table 2). The difference in clone induction between the two genotypes is a quantitative measure of recombinagenicity (9).

0 0