The work of mitochondria and chloroplasts is energy transduction in respiration and photosynthesis. The physico-chemical mechanisms of bioenergetics do not directly involve genes and heredity, and furthermore, redox chemistry is intrinsically mutagenic. Thus the small, functional genomes of mitochondria and chloroplasts are an oddity. Although extensively sequenced and catalogued, cytoplasmic genomes are still not explained. Genomic lethargy is not the answer. Some genes linger from the bacterial ancestors of these organelles, true, but most have left, and new ones arrive. There is a mounting case for a massive and indiscriminate intracellular gene transfer between organelles and the cell nucleus, with the frequency of relocation being comparable to that of mutation. Nevertheless, a few organellar proteins, all working at the core of bioenergetics, always seem to keep the genes encoding them close at hand. Stability amid flux suggests the invisible hand of selection. Selection for what? There are clues, and the beginnings of experimental support, for the theory that expression of mitochondrial and chloroplast genes is regulated by the function of their gene products. For safe and efficient energy transduction, genes in organelles are in the right place at the right time.

Two protease activities of pea chloroplasts, one located in the stroma and the other associated to the thylakoid membrane, are described. Both proteases catalyse the endo-proteolytic cleavage of a peptide corresponding to the N-terminal loop and the first turn in helix-B of light-harvesting complex II (Lhcb1 from pea). The stromal protease cleaves preferentially on the carboxy-side of glutamic acid residues. Inhibitor studies indicate that this protease is a serine-type protease. The protease was partially purified and could be correlated to a 95-kDa polypeptide band on SDS-polyacrylamide gels. The 95 kDa protein was partially sequenced and showed similarity to an to an 'unknown protein' from A. thaliana (in the NCBI public database) as well as to a glutamyl endopeptidase purified from crude extract of cucumber leaves. It is concluded that the stromal protease is a chloroplast glutamyl endopeptidase (cGEP). The protease localized in the thylakoid membrane, cleaved the peptide at only one site, close to its N terminus. The activity of the thylakoid-associated protease was found to be drastically increased in the presence of the reducing agent 1,4-dithiothreitol. Inhibitor studies suggest that this protease is a cysteine-or serine-type protease. The possible roles of these proteases in the regulation of photosynthetic electron transport and in the chloroplast homeostasis are discussed.