gocentral
commented
15 years ago Hi Val and Ivo,
This result is very strange! I was unaware of this paper and, despite a lot of reading on import, had never seen Tim23 described as spanning the outer membrane. In several recent reviews with nice graphics, it's shown as being integral to the inner membrane only. At SGD we're currently reviewing and updating annotations for import complexes (though we haven't gotten to this one yet), so I looked into this in more detail.
PMID:15797382 says "The N-terminal 50 residue segment of Tim23, which has been reported to span the outer membrane (Donzeau et al., 2000), is absent in tim23-3 mitochondria (Figure 2D) (Chacinska et al., 2003). The formation of Oxa1TOM, however, was not affected in tim23-3 mitochondria compared to wild-type mitochondria (Figure 2E, lane 8). Together with the finding that this N-terminal segment of Tim23 is not conserved in evolution (data not shown), this supports the view that the predicted outer membrane-spanning topology of Tim23 is not critical for preprotein import or accumulation (Chacinska et al., 2003)."
This article (not in PubMed): The Role of the TIM23 Complex and Its Associated Motor Complex in Mitochondrial Protein Import, The Enzymes, Volume 25, 2007, Pages 387-411 Jan Dudek, Bernard Guiard, Peter Rehling says "Interestingly, based on protease protection assays, Tim23 was suggested to span the outer mitochondrial membrane with its N-terminal amino acids [115]. However, an interaction between this domain and the TOM complex was not found and later studies showed that the N-terminus of Tim23 is dispensable for formation of a TOM–TIM supercomplex connected by a translocating precursor protein [116]."
PMID:18672008, a 2009 review, does show the N-terminal domain of Tim23 as embedded in the outer membrane (see Fig.3)., and says "The N-terminal ca. 20 amino acid residues of yeast Tim23 are apparently inserted into the outer membrane as they are accessible to protease added to intact mitochondria [35]. We have recently shown that this association is dynamic and depends on the translocation activity of the TIM23 complex [36]"
Finally, that last cited reference (PMID:18418384) suggests that Tim23 inserts the domain into the outer membrane during translocation of imported proteins, and removes it when not translocating.
My take on all this is that the outer membrane localization of that domain is somewhat controversial. Even if it's accurate, I would say this is definitely NOT a reason to give outer membrane parentage to the whole TIM23 complex - that would be misleading. I would also argue that Tim23 itself should not have an outer membrane CC annotation. There's no question that the bulk of the protein is stably integrated into the inner membrane - it forms a channel. If it only sticks its little domain into the outer membrane every once in a while...it seems to me that an outer membrane annotation would be misleading. Although we might need more discussion on that one, since we do use GO to capture transient locations (like transcription factors that shuttle between the nucleus and cytoplasm).
Maria
Original comment by: mariacostanzo
OK based on this it seems that mitochondrial inner membrane presequence translocase complex should also have mitochondrial outer membrane as a parent. Then. I will look into this
Thanks Val
"Ivo Pedruzzi via RT" <sp_upd@expasy.org> wrote: > > Hi Val, > > > > according to PMID:10830167, the S.cerevisiae ortholog spans both > > membranes (inner and outer mitochondrial membrane). We have transfered > > this annotation by similarity to the S.pombe protein. > > > > Best > > -ivo
Reported by: ValWood
Original Ticket: "geneontology/ontology-requests/5826":https://sourceforge.net/p/geneontology/ontology-requests/5826