Development of the male scent organ of creatonotos-transiens (Lepidoptera, Arctiidae) during metamorphosis

(1) The male abdominal scent organ (corema) of the arctiid moth Creatonotos transiens consists of a basal bladder and four tubes. It can be everted from the sternal intersegmental membrane 7/8. Its scent hairs (scales) produce and release the pheromone hydroxydanaidal, which attracts both sexes. Pyrrolizidine alkaloids (PA) ingested by the larva with its food are not only precursors of the pheromone but also a morphogen, which quantitatively controls the growth of the pupal corema and, thus, its final size and number of hairs. (2) The coremata arise from epidermal anlagen at the anterior border of the 8th abdominal sternite. If male larvae are fed 1 mg PA these organs begin to develop from small vesicles, and four tubes then arise during the first 3 pupal days. The corresponding mitoses reach their peak at 36 h. During the next 2 days the tubes shorten, while the walls become thin and doubly folded. The total surface of the corema increases about 20 times because of the shape transformation of the epidermal cells from prismatic to very flat. (3) The scent hairs originate from trichogen cells, which arise together with their associated tormogen cells during the 1st pupal day by way of differential mitoses. As the trichogen cells grow, their nuclei enlarge by way of endomitoses, elongate distally, and thus produce the hairs that extend into the lumen of the corema. Tormogen cells degenerate by the 8th day at latest. The hairs in each tube form a thick, caudally oriented bundle. The hair cells are finally bottle-shaped and at day 6 they extend freely into the hemolymph space. They are probably also the pheromone-producing cells in later pupal and early imaginal life. Mitoses that produce trichogen cells stop after the 1st day, those producing epithelial cells 2 days later. This delay shifts the ratio of the two cell types from about 1:11 (18 h) to 1:40. (4) The processes hitherto described refer to "normogenesis" with ample PA supply. "Control " coremata in PA-free or PA-deficient specimens develop in principle in the same way, but at a slower rate, with minimal hair cell numbers barely 1/10th of normal, or at any rates between, depending upon the earlier PA supply. The size of control coremata varies from very small to small; even the hair cells and the hairs are smaller. (5) PA regulates corema development quantitatively through the number of mitoses of its cells and of endomitotic steps of the hair cells. In PA-treated specimens the corema anlage is already advanced prior to pupation, at about the time when its sensitivity to PA influence terminates, in the early prepupa. Since PA only affects the anlagen of the corema and not that of any other body part (not even the basal coremal bladder), we postulate a selective interaction of PA with the presumptive corema cells. We found earlier that ecdysone is also involved, since the respective cell numbers can only be realized if this hormone is present