It has been previously shown that large granules of protein accumulate in the larval fat body of holometabolous insects shortly before pupation. In the butterfly Calpodes ethlius there is evidence that the protein in these granules is sequestered from the hemolymph and that their accumulation is controlled by the molting hormone, ecdysone. Evidence as to hormonal control of granule accumulation in Drosophilamelanosaster is conflicting.
In this thesis it is shown that if 3H-labelled proteins and 14C labelled amino acids are injected into D. melanogaster larvae at the time of granule formation, the proteins are incorporated into the granules but the free amino acids are not. The conclusion that all proteins in the granules are preformed hemolymph proteins is reinforced by the observation that injection of cycloheximide shortly before first appearance of the granules did not in any way interfere with their formation.
It is shown that injection of ecdysterone into early third instar larvae will induce precocious formation of granules. When fat bodies are incubated in vitro in media containing serum, addition of ecdysterone consistently causes granule formation. However, granules sometimes form in the absence of added hormone. Addition of 50 mg/ml or more protein consistently causes the formation of small granules, whether or not hormone is also present.
When fat bodies are incubated in media without serum, granules consistently appear in aerated media without addition of hormone. However, addition of hormone causes the formation of larger granules which appear earlier.
It is therefore concluded that in D. melanogaster ecdysone is sufficient but not necessary for granule formation.
Since the granules appear seventeen hours before pupariation, while ecdysone is secreted only four hours before pupariation, it is suggested that ecdysone is not involved in control of the natural appearance of the granules in D. melanogaster. The hypothesis is presented that the granules form spontaneously as soon as the juvenile hormone titer falls below a certain level.
In the course of this work it was incidentally noticed that injection of ecdysterone into early third instar larvae caused a syndrome which resembled a partial and abortive prepupal molt. No puparium was formed and the larvae died within a few days.
