Putative channel-forming components

While Toc75 has always been the best candidate for the central channel-forming com-ponent of the outer envelope membrane in higher plants [136], the identity of the inner envelope membrane pore is still unclear. There are some candidates, which can fulfil such a function: Tic20 [31], Tic21 [149] and Tic110 [60].

Tic110 was the first Tic component to be found in vascular plants. Firstly it was described as a component of the late translocation intermediate [136, 159]. It is postulated that Tic110 is an integral membrane protein possessing two hydrophobic transmembrane α-helices at the N-terminus [77]. Sequence analyses of Tic110 in P. tricornutum resulted in one to two possible transmembrane domains, depending on the used program (Table A.2). The role and the localisation of the hydrophilic rest of the protein is still unclear. According to one theory the C-terminal part is localised in the stroma, and probably functions to recruit chaperones [64]. In chloroplast import the relevant chaperone is Hsp93 (Heat shock protein 93kDa) [117] which we also could identify in theP. tricornutumgenome (PtHsp93 Protein ID: 25856). To check whether PtTic110 is integrated to the innermost envelope membrane we made a carbonate extraction (Figure 2.8D). Only in the integral membrane fraction we get a signal with the right size, indicating that PtTic110 is integrated in a membrane. This result fits also with the findings, that P. tricornutum cells expressing PtTic110full:GFP shows GFP fluorescence inside the plastid (Figure 2.21C).

2.5.Discussion

et al. (1997); Küchler et al. (2002); Heins et al. (2002); tested additionally with self-assembling GFP; C-terminus faces presumably into the ies, because of the interaction between PtTic21 with PtTic22; ^a,b,cProtein IDs refer to the following databases: ^aP. tricornutum v2.0 (http://genome.jgi-psf.org/Phatr2/Phatr2.home.html);^bT. pseudonana v3.0 (http://genome.jgi-psf.org/Thaps3/Thaps3.home.html);^chttp://www.ncbi.nlm.nih.gov; n.i.:

not identified; n.e.: no evidence; n.d.: not detected; n.dtm.: not determined; Pt/Tp: Phaeodactylum tricornutum/Thalassiosira pseudonana

Protein Identified diatom

Tic20 56631^a / 270420^b Channel component yes n.e. n.e. n.e. pl integral stroma / stroma

Tic21 XP_002183847.1^c / 270422^b Channel component 

Iron permease no n.e. with PtTic22 n.e. pl unclear stroma / ies^□

Tic22 EEC47673^c / 4604^b Scaffold protein between 

Toc and Tic  n.e. n.e. with PtTic21 n.e. ies* unclear n.e.

Tic32 11808^a / XP_002288562.1^c Redox regulator n.e. CaM binding

NAD binding n.d. n.d. pl peripheral n.e.

Tic40 n.i. Chaperone binding − − − − − − −

Tic55 50613^a / 270428^b Redox regulator unclear

Fe2+ binding 2Fe‐2S binding

PAO C‐x‐x‐C

n.d. n.d. pl integral n.dtm.

Tic62 35796^a / XP_002295095.1^c Redox regulator n.e. NAD binding n.d. n.d. pl soluble n.e.

Tic110 50540^a / XP_002287032.1^c Channel component unclear − n.d. n.d. pl integral n.dtm.

predicted protein domains Interaction (using split GFP) Localisation

Alignments with Tic110 proteins from different organisms revealed a leucine-zipper-like motif (L-x-x-L-x-x-x-L-G) within the C-terminal half. The functional relevance of this motif is not clear, even though a A. thaliana mutant lacking the C-terminal part has been shown to be disturbed in plastid protein import [64, 73].

Another candidate for a channel-forming component, which is inserted into the inner envelope membrane via four predicted transmembrane domains is Tic20 [84].

There are some structural relations between Tic20 and the mitochondrial inner mem-brane translocon components Tim17 and Tim23 [73]. These Tims also possess four transmembrane helices and form an import channel in the innermost mitochondrial membrane. Sequence analyses on PtTic20 show structural similarities to these Tims but also to higher plant Tics. They possess also three to four predicted transmembrane helices. Expression of the PtTic20full:GFP in P. tricornutum led to GFP accumula-tion within and around the plastid silhouette, probably composed of membrane bound fusion protein aggregates (Figure 2.15). The presence of a functional stroma targeting sequence (Table A.1) together with the finding that plastid silhouettes are labeled in full length GFP fusions, suggests that PtTic20 inserts into the innermost plastid mem-brane via the plastid stroma, where it might operate in aggregates. Our results from the localisation studies using carbonate extraction confirm these findings. Western blot analyses made with fractions from carbonate extraction show a band with the right size in the integral membrane fraction beside some further signals (Figure 2.8A;

marked with an arrow) which we also can observe in the control (Figure 2.9A and B; marked with a star). For this reason it is thinkable that this might be unspecific signals. Because of the experimental procedure, we could not determine the protein concentration. Therefore it is possible that we loaded more from the control than from the Tic samples on the SDS-gel, which finally causes these unspecific bands. This also could be the case for the following Tics (PtTic32, PtTic55 and PtTic62; Figure 2.8 and Figure 2.9) we tested. Further results we observed in orientation studies using self-assembling GFP, support the role of PtTic20 as an integral membrane protein localised in the innermost envelope membrane (Figure 2.5). Here we co-transformed PtTic20Cfull:GFP11 and PtTic20N:GFP11, respectively, with the stromal marker fusion protein PtOEE1pre:GFP1-10 and the interenvelope space marker PtPlSP1 (Pt-PlSP1 is localised in higher plants in the thylakoids where it processes signal peptides of thylakoid targeted proteins, but also in the interenvelope space where it is involved in processing of Toc75 [65, 20]). GFP fluorescence was only detected within the plas-tid, using the combination with the stromal marker protein, suggesting that the

C-and N-terminus faces into the stroma, as van Dooren et al. demonstrated in Toxo-plasma gondii (Protein ID: ACB58125.1) [152]. Here it was again not possible to get a distinct prediction of transmembrane domain number, it ranges between zero and four predicted transmembrane domains (data not shown).

Tic21 is an alternate putative candidate for the inner-membrane translocation chan-nel. Reminding of Tic20, it is an integral membrane protein with four predicted trans-membrane domains. Interestingly, attic21 null mutants show a sickly phenotype in A. thaliana and accumulate preproteins in the cytosol [149]. These findings indicate a defect in protein import. But the same protein was also characterised in Duy et al.

[38] as a metal permease. P. tricornutum transformants expressing PtTic21full:GFP showed a GFP fluorescence within the plastid and accumulation of GFP in dots close to the plastid (Figures 2.16). The presence of a functional stroma targeting sequence together with the finding that GFP accumulates in the plastid and partially in dots, indicates that PtTic21 could insert into the innermost plastid membrane via the plastid stroma as PtTic20. In order to check whether N-terminus and C-terminus, face into the stroma or the interenvelope space, we used the self-assembling GFP system. Only in P. tricornutum cells expressing PtOEE1pre:GFP1-10 together with PtTic21Cfull:GFP11 we could observe a GFP fluorescence within the plastid, demon-strating that the N-terminus faces into the stroma. In contrast, no fluorescence could be detected when we co-transformed PtPlSP1full:GFP1-10 together with PtTic21C-GFP11 and PtPlSP1full:GFP1-10 and PtOEE1:GFP1-10, with PtTic21N-PtTic21C-GFP11. To investigate protein-protein interactions in vivo we used the split-GFP system. Here we observed a GFP fluorescence when we co-transformed PtTic21full:NGFP with PtTic22full:CGFP, which indicates an interaction between these two proteins. The finding that the N-terminus faces into the stroma and we get an interaction between PtTic21 and PtTic22 (Figure 2.11C), suggests that the C-terminus faces into the interenvelope space, and for this reason an additional role of PtTic21 as an integral membrane protein, which is localised in the innermost membrane is proposed. The role Tic21 finally may have is still unclear. But the results show that PtTic21 possi-bly may be part of the import machinery. Phylogenetic analyses of the three putative channel forming proteins show that Tic20, Tic21 and Tic110 form in each case two distinct clades: a red clade and a green clade (highlighted in red and green). The red clade of Tic20 contains the cyanobacterial homologs, as well as a subclade that contains Tic20-like proteins. In general, the diatom Tic20, Tic21 and Tic110 branch together with other secondary plastid containing algae, from the red lineage (with

a representative from the red algae: C. merolae) rather than with primary algae (containing primary plastids). This fits with the hypothesis that plastids of chroma-lveolates share an evolutionary common origin [46]. For the remaining Tics, Tic32, Tic55 and Tic62 it was not possible to create a phylogenetic tree, because not enough sequences are available. Until now there are diverse candidates described that could build an import channel in the innermost envelope membrane in plants. But it is still not clear whether they are part of a Tic complex or they function independently of each other.