We are interested in a small protein from Arabidopsis that we first thought
to be secreted because of its putative 22 aa pre-senquence with secretion
signal characteristics: a basic amino-terminal end, a hydrophobic central
region, a polar C-terminus and a predicted cleavage site consistent with
the (-3;-1) rule proposed by von Heinje.
However, we unambiguously immunolocalize this protein in the chloroplast,
possibly in the thylakoid lumen (immunogold at E.M.level not yet done).
Is anybody aware of a nuclear encoded chloroplast protein, devoid of
N-terminal envelope transfer domain, eventually ending up in the thylakoid
lumen due to its putative lumen translocating domain such as the one
described above (clearly not bipartite).
Any suggestion is wellcome. Many thanks.
herzog at ccomm.grenet.fr