Résumé
The endogenous retrovirus gypsy is controlled by the Drosophila gene flamenco (flam). New insertions of gypsy occur in any individual Drosophila if its mother is homozygous for the flam1 permissive allele and contains functional gypsy proviruses. The ovaries of flam1 females also contain high amounts of gypsy RNAs. Unexpectedly however, gypsy derepression does not occur in the flam1 female germ-line proper but in the somatic follicular epithelium of the ovary. Since extracts from these females are able to efficiently infect the germ-line of a strain devoid of active gypsy proviruses, we assume that a similar kind of germ-line infection, which would occur inside the flam1 females themselves, could be required for gypsy insertions to occur in their progeny. This hypothesis was confirmed by electron microscopy observations showing that non-enveloped intracytoplasmic particles containing gypsy RNAs accumulate in the apical region of the flam1 follicle cells, close to specific membrane domains to which the gypsy envelope proteins are targeted, whereas both are absent in the flam+ controls. Low amounts of similar virus-like particles were also observed in flam1 oocytes, but it is not yet known whether they entered passively or as a result of membrane fusion. This is the first report of the beginning of a retrovirus cycle in invertebrates and these observations should be taken into account when explaining the maternal effect of the flamenco gene on the multiplication of gypsy proviruses.The endogenous retrovirus gypsy is controlled by the Drosophila gene flamenco (flam). New insertions of gypsy occur in any individual Drosophila if its mother is homozygous for the flam1 permissive allele and contains functional gypsy proviruses. The ovaries of flam1 females also contain high amounts of gypsy RNAs. Unexpectedly however, gypsy derepression does not occur in the flam1 female germ-line proper but in the somatic follicular epithelium of the ovary. Since extracts from these females are able to efficiently infect the germ-line of a strain devoid of active gypsy proviruses, we assume that a similar kind of germ-line infection, which would occur inside the flam1 females themselves, could be required for gypsy insertions to occur in their progeny. This hypothesis was confirmed by electron microscopy observations showing that non-enveloped intracytoplasmic particles containing gypsy RNAs accumulate in the apical region of the flam1 follicle cells, close to specific membrane domains to which the gypsy envelope proteins are targeted, whereas both are absent in the flam+ controls. Low amounts of similar virus-like particles were also observed in flam1 oocytes, but it is not yet known whether they entered passively or as a result of membrane fusion. This is the first report of the beginning of a retrovirus cycle in invertebrates and these observations should be taken into account when explaining the maternal effect of the flamenco gene on the multiplication of gypsy proviruses.