Non-male-killing spiroplasmas in Drosophila hydeiT. OTA, M. KAWABE, K. OISHI, AND D. F. POULSON M ATERNALLY transmitted microorganisms that cause male-specific lethality in progeny of Drosophila have been found in some flies from natural populations of four neotropical species all belonging to the willistoni species group (subgenus Sophophora). These microorganisms, previously called SR-spirochetes, are now recognized as spiroplasmas: helical, motile, cell wallfree prokaryotes in the class Mollicutes 1 -. The abnormal sex-ratio condition due to these SR-spiroplasmas has been studied with two aspects in mind: that it might provide a means to investigate the process of sex-differentiation in the host Drosophila 1 -2 -7 -8 and to learn more about the basic biology of spiroplasmas 3 ' 6 -9 " 11 ' 13 .It has been proposed that SR-spiroplasmas produce an exotoxin-like substance, androcidin, that is responsible for the lethality in male zygotes 2 . In spite of a rather extensive effort to recover non-male-killing variants of SR-spiroplasmas from natural populations of the willistoni species group flies none has been found (Poulson, unpublished). This along with our present inability to grow SR-spiroplasmas in vitro has prohibited us from attempting studies on the genetic basis of the male-killing effects of SR-spiroplasmas. We report here a new occurrence of spiroplasmas in a natural population of Drosophila hydei (repleta species group, subgenus Drosophila) which have no male-killing effect in host flies.

A sex-transformation gene, tra-2OTF, of Drosophila melanogaster is described. The gene is an allele of tra-2 (transformer-2, 7--70+/-), acts only on XX individuals and transforms them into morphologically nearly normal but sterile males. Neither a male-specific lethal gene mle nor the male-killing maternally transmitted infectious SR-agent have any effect on the tra-2OTF transformed flies. External as well as internal reproductive organs of tra-2OTF transformed flies, except for the gonads, are male. The gonads, which are yellowish-colored rudimentary testis-like structures in tra-2 transformed flies, are rudimentary ovaries without yellowish coloration that sometimes develop to near maturity. Numbers of bristles in each of the female pattern in tra-2OTF and tra-2OTF/tra-2 transformed flies, while these bristles are of the male pattern in tra-2 transformed flies. Furthermore, the sixth abdominal sternite, which in normal males as well as in tra-2 transformed males lacks bristles, bears numbers of bristles in tra-2OTF and tra-2OTF/tra-2 transformed flies.

In Drosophila, vitellogenins (yolk protein precursors) are synthesized by the female fat body, secreted into the hemolymph and subsequently taken up by the developing oocytes. The male fat body, on the other hand, does not do this even when immature ovaries are transplanted into the body cavity and grow. Thus, the hemolymph vitellogenins serve as an easily detectable sexually dimorphic biochemical marker.——We have examined hemolymph vitellcgenins by SDS polyacrylamide gel electrophoresis in flies carrying various sex-transformation mutants (dsx, tra, tra-2 and tra-20TF) singly and in all possible combinations. Chromosomal females homozygous for tra or tra-2 have no detectable hemolymph vitellogenins, while those homozygous for tra-20TR exhibit appreciable levels of these proteins. Flies homozygous for dsz, bothX / X and X / Y, have hemolymph vitellogenins, although the amount is consistently smaller in the latter. Indeed, X / Y; dsx/dsx is the only genotype in which hemolymph vitellogenins are detected in the X/Y flies. A clear hierarchy of epistasis exists among these sex-transformation mutants when they are examined in various combinations: dsx > tra, tra-2 > tra-20TF. Moreover, an interaction between tra-20TF and tra was seen in these experiments: X / X; tra-20TF/tra-20TF flies show the presence of only a trace of hemolymph vitellogenins when they are made heterozygous for tra. These results, combined with observations on gonad morphology, are discussed with respect to the BAKERa nd RIDGE (1980) hypothesis of sex determination.

The homoeotic mutations of Drosophila melanogaster, '` podoptera" and "tetraltera", were studied in detail by R. B. Goldschmidt and his colleague (Goldschmidt 1945(Goldschmidt , 1952Goldschmidt, Hannah and Piternick 1951;Villee 1942). These mutations affect the wings in such a way as to transform them into halters or leg-like structures. Goldschmidt said that "podoptera types are not at all rare but are practically ubiquitous in Drosophila stocks. One of the reasons that their hereditary nature was overlooked in all previous years of Drosophila work is that their penetrance is normally very low, particularly in mass breeding". The same is true of Drosophila virilis.We frequently found flies which lacked wings in stocks of D, virilis, but we could not study the hereditary nature of this type of abnormality because of its extremely low penetrance. Recently, we obtained two tetraltera-type strains which we could study morphologically as well as genetically. MATERIALS AND METHODSIn the course of studies on melanotic tumours formed on the body surface of old imago of D. virilis, T. Kanehisa in our laboratory found flies lacking one wing in the wild strain collected at Sapporo (Hokkaido, Japan). This new mutation had various characteristics. These were, transformation of the wing into a halter-or palp-like structure with or without hairy warts and knobs, and sometimes reduction or defects in the dorsal region of the thorax. The extent of the abnormalities varies from individual to individual and is for the most part asymmetric. As the mutation phenotypically resembles `~tetralteraBd" of D. melanogaster, we named this tetraltera (tet). Mutant stock with a stable penetrance of about 10 percent was obtained by mating a virgin mutant type female with a mutant type male. Another mutation of the same type was found in the stock Notch3a of D. virilis. This mutant strain had so low penetrance that selections had to be repeated several times to obtain a stock with a stable penetrance of about 0.07 percent. Linkage relations were studied by the usual method. Flies carrying recessive marker (r) were mated with tet flies, and F, wild-type flies were inbred. When a dominant marker (D) was used, males hetero-