The synthesis of storage proteins in ripening Agrostemma githago seeds was studied by in vivo pulse and pulse-chase experiments with labeled amino acids and labeled glucosamine. It was found that storage proteins were not synthesized directly, but via cleavage of several lrge precursor proteins. Two disulfide-linked proteins of38 and 25 kilodaltons were synthesized via a single large precursor protein. This precwusor protein contained internal disulfide bridges, at least one of which is involved in holding the subunit structre together following cleavage of the precursor. A similar mode of biosynthesis was noted for two other disulfide-linked proteins of 36 and 22 kilodaltons. The half-life of the precursors was about 2 hours. This mode of processing is a to the synthesis of legumin in legumes and globulin in oats. A third pair of disulfide-bonded proteins (41 and 23 kilodaltons) was synthesized from a precursor protein in several steps. These included a legumin-lke cleavage, whereafter the subunits remained disu}fide-bonded. Then, from the largest subunit, a part was cleaved off, probably a storage protein of 17 kilodaltons. This 17-kilUlton protein was not disulie-bonded to the 41 and 23-kihlodton complex. The first processing step was fast, the second slow: The half-lives of the precursors were about 3 and 10 hours, respectvely. Finally, a group of 16-and 17-kilodalton proteins was synthesized by cleavage of arge precursor proteins, likely in two steps. After cleavage, the proteins were not disulfide-bonded. The half-lfe of the precursors was short, less than 1 hour. In addition, for the 38-, 23-, and one of the 17-kilodalton proteins, a small decrease of relative molecular weight was observed as a last processing step. This was likely due to deglycosylation.Numerous studies have been carried out on the synthesis of storage proteins in ripening seeds (for reviews, see 6, 23). It has been found that in many species storage proteins are not synthesized directly but by cleavage of high mol wt precursor proteins. The precursor proteins may undergo several co-and post-translational modifications before yielding the final storage proteins. First, comparison of in vivo and in vitro translation products showed that the latter are normally a few kD larger. Probably, in vivo, a signal peptide is removed co-translationally (7,19). Second, some storage proteins are glycosylated. It was shown for phaseolin in beans (3) and conglycinin in soybeans (24) that cotranslational glycosylation is followed by a second glycosylation step. Third, following synthesis, the precursor may undergo cleavage. It was found for legumin in peas (1 1, 25), glycinin in soybeans (1, 24, 27), and globulin in oats (5) that the precursor protein is split into two subunits which, after cleavage, remain connected by a disulfide bridge. It might be that, such as in the processing of pro-insulin in animals, a connecting peptide is excised (26). In peas the lower mol wt vicilin subunits are also derived from large precursor proteins. After the s...