From the information collected based on sequence and structural database and literature searches, we found that there are two possibilities of what c1orf41 function could be. Firstly, it might be a heat shock protein and secondly it might have a function that is related to galactose binding or cell adhesion.
Heat Shock Protein
Based on Blast, it was shown that this protein has high similarity sequence with small heat shock proteins (sHsp) from various eukaryotes. Our protein was a mutant of sHsp family B member 11 where aspartic acid at position 109 has been deleted. The first ten residues were believed not to be part of the protein but most likely an N-terminal histidine tag used during purification of the protein. This was proven by the multiple sequence alignment as these residues does not align with other organisms. The sequence for alpha-B-crystallin domain found in sHsp is also present in c1orf41. However, we were unable to find any structure for alpha B crystallin so we could not compare the structural similarity with our protein. But our protein has a mainly beta-sheet structure which is observed in alpha crystallin domain. Therefore, there is a possibility that c1orf41 may be a sHsp.
sHsps are molecular chaperones that can be found in various organisms, however their sequences are poorly conserved. Studies have shown that the number of gene encoding sHsp increases in higher eukaryote. This may be due to the specific function and localization of sHsp (Haslbeck et al., 2005). Since it was shown to be highly expressed in cancer cells, the function of this protein may be related to cancer. The loss of Asp109 may lead to cancer or alternatively, cancer cells may have produced this mutated protein.
The Gene Ontology (GO) suggested that c1orf41 is likely to be an intracellular protein and this agrees with the findings by Bellyei et al(2007). Many apoptotic key molecules (such as cytochrome c and pro-caspase 3) are located within mitochondria. Studies have also shown that many cellular proteins that have anti-apoptotic function are located within mitochondria in order to prevent the release of pro-apoptotic proteins (Concannon & Samali, 2003). In another study, Kang et al. (2004) suggested that in normal conditions, sHsps are located in cytoplasm, near the golgi complex and will move to the nucleus when cell is induced by stress. Therefore, since c1orf41 is an intracellular protein, this characteristic is typical of sHsp which are found mainly in the cells. GO also suggested that this protein is probably involved with cellular response to stress as obeserved in sHsps.
Since c1orf41 showed high homology with sHsp 16.2 (85% identity and 95% similarity) (Bellyei et al, 2007), it is possible that it has similar physiological role to sHsp16.2. Bellyei et al. (2007) reported that sHsp 16.2 is always bound to Hsp 90 which mediates its function. It inhibits cell death by stabilizing the mitochondrial system (inhibiting cytochrome c and caspase 3), increasing the lipid raft formation and increasing Akt phosphorylation.
Akt is a member of protein kinase B which plays an important role in cellular signaling (cytoprotective protein kinase). Akt inactivates tumor suppressor through a phosphorylation-dependent process (Bellyei et al, 2007). C1orf41 structure showed that it has a loop that binds to Ca2+ ion. The presence of calcium is presumably related to the Akt phosphorylation function. Sriram et al. (1997) reported that the presence of Ca2+ metal ions support the phosphorylation of Hsp70. Moreover, the protein-protein interaction (String result) showed that C1orf41 has a linked function to RAYL which is a member of RAS-like-GTPase superfamily. It was reported by Yue et al (2003), the overexpression of RasGAP (Ras GTPase-activating protein),negative regulator of Ras, will enhance the phosphorylation and activity of Akt. Ras regulated intracellular signaling network, such as proliferation, differentiation, cell adhesion, apoptosis and cell migration. Mutation of Ras gene leads to tumor formation as it keeps the cell in the proliferative mode (Karp, 1999).
DALI result showed that c1orf41 has high structural similarities to sialidases and galactose oxidases. However the proteins show minimal sequence similarity with c1orf41. A search using Pfam indicated that this protein has a F5/8 type C domain which is a member of the discoidin domain family. The Pfam search also showed that this domain is also observed in proteins from the DALI result. We concluded that the DALI result obtained was mostly based on the presence of F5/8 type C domain in c1orf41 and the other proteins.
The discoidin protein is also known as the galactose binding domain-like. Members of this family such as bacterial sialidase and galactose oxidase has a galactose-binding characteristic. It is also known that proteins with F5/8 type C domain are involved in cell adhesion. Cell adhesion is another biological process that was also stated in the GO result. Therefore, we hypothesized that this protein may have a galactose-binding property and possibly cell adhesion function.
Sujatha and Balaji (2004) have reported that the residues and position of galactose binding site is conserved within family but not conserved between families. From previous literature H-(2x)-W-(29x)-R-(2x)-S-(2x)-E are the residues important for galactose binding within this family, however these residues were absent in c1orf41 (Gaskell et al., 1995). Even though the structure showed similarity with galactose binding domain, the absence of galactose binding site indicates that this protein does not bind to galactose.
F5/8 type C domain is involved in cell-cell interaction and cell recognition. Highly expressed, the protein can lead to breast carcinoma. Northern blot analysis of several human breast carcinoma cell lines (T-47D, BT-20, and MCF-7) showed that discoidin domain transcript is abundant in these cell lines (Johnson et al, 1993). Apart from that, this domain is also involved in cell adhesion. However, due to the lack of the RGD motif that is present in cell adhesion proteins, we deduced that this protein is not involved in cell adhesion (Bianchet et al., 2002).
As described in the Results section, c1orf41 has a beta barrel structure where the beta strands are organized to form a barrel structure. We found that c1orf41 has a jelly-roll fold that is characteristic of a discoidin domain member (Hidai et al., 2007). Description from Pfam indicated that this domain has two conserved alpha helices which are also present in c1orf41. These alpha helices are known to interact with cell membrane. However we found the residues on c1orf41 alpha helices to be mainly charged, therefore it is less likely to bind to the membrane. Also, many proteins with this domain have disuplhide bonds. This is not observed on our protein. In addition to that, members of the discodin family are generally extracellular or membrane proteins. Previous studies have reported that c1orf41 was found in the cytoplasm and the surface of the protein are mostly charged, therefore we do not think that it is a membrane protein.