Identification of a gene in Synechocystis PCC6803 with similarity to pea Tic22 indicated a cyanobacterial origin of the protein. More recently, a Tic22 homologue in Anabaena sp. PCC 7120 was studied and shown to be essential for development. This protein localized in the thylakoids and the periplasm, and could be functionally replaced with a plant orthologue. Immunoprecipitation after chemical cross-linking revealed a physical interaction with the outer envelope biogenesis factor, Omp85, suggesting a function of cyanobacterial Tic22 in outer membrane biogenesis. Moreover, three-dimensional structure analysis of the Anabaena protein identified conserved hydrophobic pockets similar to those of ClpS or BamB, suggesting a possible chaperone function. A Tic22 homologue was also identified in Plasmodium falciparum, and found to be peripherally associated with apicoplast membranes that are analogous to the chloroplast inner envelope membrane. Subsequently, a Tic22 homologue in a similar parasite, Toxoplasma gondii, was shown to be apicoplast-localized and crucial for both parasite survival and protein import into the apicoplast stroma. Structural analysis of P. falciparum Tic22 revealed a fold conserved from cyanobacteria to plants, incorporating non-polar grooves on each side of the molecule. Moreover, these grooves allow the apicoplast protein to function as a chaperone. Such a chaperone had not previously been known to exist in the intermembrane space of plastids. In Arabidopsis, two Tic22-related genes have been identified, and these are termed atTIC22-III and atTIC22-IV according to the chromosomal location of the corresponding genes. In this paper, we present data on the evolutionary, molecular and genetic analysis of these homologues. Very recently, another research group independently reported on the analysis of these genes, and so we relate our findings to those derived from that study. A keyword search in the Aramemnon database identifies a third potential member of the Tic22 gene family in Arabidopsis. Attempts to include its full-length sequence in our analysis resulted in poor alignment results, as its overall similarity to atTic22-III and atTic22-IV is very low; e.g., its predicted mature domain shares just,21�C23% amino acid sequence identity with those of the other Arabidopsis sequences or pea Tic22. Moreover, only part of this sequence is homologous to the other Arabidopsis sequences. Putative orthologues of At5g62650 were also found in other land plant species, and the aligning regions of these were included in the analysis. Figure 1 shows that these sequences form a clade after the split between red algae and other archaeplastida, but before the split between green algae and land plants. The analysis also shows that these At5g62650-related sequences have evolved at a faster rate than the Tic22 sequences, implying that they functionally diverged during evolution in similar fashion to Toc75 and OEP80 sequences. In combination, these various observations suggest that At5g62650 is not a canonical Tic22 protein and so it was not pursued further in this study. To provide experimental support for the TargetP predictions, we attempted to assess the subcellular Axitinib localization of the Arabidopsis Tic22 proteins by the analysis of YFP fusion proteins. To this end, full-length coding sequences, and two truncated sequences for each one, were inserted into the AbMole BioScience Life Science Reagents p2GWY7 vector which adds a C-terminal YFP tag. However, repeated analyses of each fusion in transfected Arabidopsis protoplasts by fluorescence microscopy failed to provide clear evidence of chloroplast localization. Although some of the YFP signal may have been associated with the chloroplast envelope, the fluorescence patterns observed were generally consistent with cytosolic localization.