Concerning the Healing of Chromosome Ends Produced by Breakage in Drosophila melanogaster

The numerous earlier cases in Drosophila which had been thought to illustrate changes in gene polarity-i.e. the conversion of interstitial genes to telomeres ("healing") and vice versa, and occasionally, the formation of branched connections-have on more searching analysis given evidence of involving, instead, only the union of the broken ends of chromosomes in a new arrangement. Moreover, several large-scale earlier experiments, designed to test the occurrence of healing by means of quantitative studies of the effects of irradiation of spermatozoa, have failed to give evidence of healing. Such studies have included attempts to obtain healing both in eu- and in hetero-chromatin, when the broken ends had resulted from X-rays, from ultraviolet, or from the mechanical rupture of dicentric bridges. It is true that, since the resolution of most of the earlier moot cases, some additional cytological evidence suggesting the conversion of interstitial genes to telomeres and vice versa in Drosophila has been reported, but these later cases have not been reinvestigated in the light of modern knowledge of the possibilities, and they must be regarded as far from conclusive. New attempts are herein reported to obtain evidence regarding the healing of breakages in chromosomes derived from spermatozoa exposed to some 4000 r of X-rays. Two types of X-chromosomes, especially suitable for the detection of such changes, were used, namely Bar-M1 and the synthetic (crossover) chromosome designated as y·Dp(scV1), explained in the text. Only 5 cases which might have represented healed breaks, as compared with 18 identified multi-break cases, involving changes in the connections of broken ends, were obtained. This frequency for the cases of possible healing is too low to afford evidence in favor of this phenomenon, in view of the fact that in our experiment a number of types of interchange of connections of broken ends, which collectively could well have been as frequent as this, would have produced individuals of the given phenotypes, identical with those which would have been produced by healed breakage. Pursuant of the articles by Rapoport (1940, 1941, but only recently available to us), reporting the formation of detached X-chromosomes as a result of the healing of breakages when Drosophila females with attached-X's and no Y were X-rayed, a similar study has been carried out by us. Special attached-X's were constructed, of a type designated as "snoc" (see text), which were particularly suitable for the finding of such cases. Females with these X's and no Y were exposed to some 2000r of X-rays. The data obtained in the F1 completely confirm Rapoport's empirical findings. However, genetic analysis of 51 of the exceptional F1 females, containing detached X's, proved that in at least 6 cases these X's had resulted from paracentric interstitial deletions. It is shown that this number of deletions is certainly a minimum one, and that the frequency of the intra-X interchanges in this material is relatively high, in view of the dearth of cases of all types of interchanges hitherto reported as having been obtained from females. Even more surprising is the consistency of the data with the interpretation that all the exceptional cases, in both Rapoport's and the present study, resulted from interchanges of connections of broken ends, rather than healing. For the consideration must not be neglected that when (as was the case for most of the exceptions here produced) oocytes are irradiated, the broken proximal portion of the X is not restricted to becoming joined with a subterminal portion broken off from the chromatid of an X chromosome, so as to give a deletion type exception, but that it may become joined, instead, with a subterminal region broken off of any autosomal chromatid. This would result in a translocational-type exception phenotypically like that expected from healing. Accordingly, the results from X-rayed females can no longer be considered as evidence for the healing of breakages with formation of new telomeres. At the same time, on the interpretation that the detachments obtained do represent interchanges of broken ends, it must be concluded that the frequency of this phenomenon in oocytes is much higher than had been suspected. Moreover, the data indicate that irradiated oocytes yield many more of these gross rearrangements than do irradiated oogonia. It is concluded that, under the wide variety of conditions in which healing of breakages has been sought in Drosophila it either does not occur or else occurs with viable products so rarely as to escape detection by quantitative means.