The main objective of this project was to identify precisely the selenoproteins and selenoprotein machinery required for their synthesis present in Colinus virginianus' genome. As explained before, Gallus gallus was chosen due to phylogenetic proximity and good characterization of its genome, but Homo sapiens and Meleagris gallopavo were used in certain cases.
A total of 34 proteins were studied, 28 from the chicken, 4 from the turkey (SelP1 and SelP2, GPx2 and SelM) and 3 from the human (SelW, GPx1 and GPx4), which were neither in turkey nor chicken but were supposed to be according to Li et al. 2018. Additionally, two more proteins that had not been found in Colinus using chicken as a query were repeated using turkey's as queries (SelH, GPx8), and in some cases, using human.
The following characterization was obtained:
- Selenoproteins: DIO1, DIO2, DIO3, GPx1, GPx2, GPx3, MSRB1, Sel15, SelH, SelI, SelK, SelM*, SelN, SelO, SelS, SelT, SelU1, TXNRD1, TXNRD2, TXNRD3
- Cys-containing homologs: GPx7, MsrA, MsrB3
- Selenoprotein machinery: eEFsec, PSTK, SBP2, SECp43, SecS, SEPHS
- Not predicted proteins: GPx4, GPx8, SelP1 and SelP2, SelW
*SelM is supposed to be a selenoprotein, however, in the reference sequence there was no Sec nor Cys residue. Despite of that, the sequence was correctly aligned, so maybe there is a part of the protein that was not well noted in SelenoDB.
From the obtained results, we can see that 20 of the selenoproteins were conserved in Colinus virginianus and we succeeded to identify and characterize them. Additionally, 3 more Cys-containing homolog proteins were predicted and characterized, as well as 6 machinery proteins.
We did fail, however, to find 5 proteins that were supposed to exist phylogenetically.
SelenoP is supposed to have 2 variants in Gallus gallus: one with many Sec residues present, which accounts for 50% of Selenium in plasma, SelenoP1, and another containing one Sec only, Selenop2, whose function remains unknown [Li S et al, 2018]. SelP2 is present in Gallus gallus but seems absent in Colinus. Firstly, we analysed it and the Tblastn obtained two hits in the scaffold VONY01000009.1 that fulfilled our criteria, and even though exonerate predicted a 2 exon gene in the reverse strand. Tcoffee could not obtain a proper alignment so we decided to do it again using another as a query the SelenoP of another bird noted in SelenoDB, Meleagris gallopavo (turkey), but it did not even succeed to get a hit with Tblastn with an identity over 70%, and neither those under 70% produced Tcoffee's alignments. One SECIS element was predicted in a possible position, but since nor us nor Seblastian got a prediction, we can say that this protein was either lost in Colinus or changed too much to be found, which is not very likely since chicken and turkey are quite near Colinus. SelenoP1 was not present in Gallus gallus’ SelenoDB, so we took it from Meleagris gallopavo; again, any hit accomplished our criteria. It is very strange that these two proteins are absent in Colinus virginianus. One reason could be that, according to Labunskyy et al., the SelP protein family has recently evolved, and SelP homologs are found predominantly in vertebrates, so maybe we cannot find it because they are too different in Colinus and the other studied organisms. Further research could try to find SelenoP genes in other avian species and compare them with Colinus and try to find them.
GPx8 was also lost from Gallus to Colinus, or changed too much to be detected. We tried to find it using Homo sapiens as a query, but not even a hit accomplished our criteria. Further details about the implications of this absence are explained hereunder.
According to Li et al. 2018, GPx4 and SelenoW are also present in the Gallus gallus selenogenome, but they were not present in SelenoDB, neither in Gallus gallus nor in Meleagris gallopavo. Therefore, we used Homo sapiens proteins to check if they were present in Colinus virginianus. However, neither of them gave a proper hit according to our criteria. Due to this, we can either think that none of them is present in Colinus genome, or that they have evolved so differently in human and Colinus that we are not able to find ressemblances anymore, even though we did not encounter this problem with GPx1, that was also obtained using a human protein query. It would be necessary to repeat the process using chicken proteins as a query, in order to check if they are conserved as in chicken or not.
GPx8 and GPx7 are duplications from GPx4. Since the three of them have the same functions, it could be possible that GPx4 and GPx8 were lost in Colinus, but, as commented above, further research would be needed in order to double check these results.
SelW functions remain unknown, although it has been speculated to help redox reactions in heart muscle. Since it is not confirmed, we cannot be sure of the implications of its absence.
Most of the predicted proteins were very well characterized. Some of them, however, did not begin with a M, so further research would be needed on this issue. The problem lays most likely on the proteins from SelenoDB used as queries, since they did not have M at the beginning neither, so a better characterization of them would be needed.
Another encountered problem was the lack of prediction of Selenoproteins and SECIS elements by Seblastian and SECISearch3 where we had found one. We decided to trust our results, since their website uses databases of different species to do automatic predictions based on algorithms, and this could lead to not predicting a SECIS or a Selenoprotein even if it is there. Therefore, we consider protein conserved if we obtain a good Tcoffee alignment, with Seblastian and SECISearch3 being useful to corroborate our results.
All in all, we consider that a high number of proteins have been identified and characterised successfully in Colinus virginianus, providing a small contribution to the current knowledge of selenogenome, which should continue to be enlarged.
Additionally, we would like to mention that, in order to continue with the research in selenoproteins and their machinery, the genome of more bird species should be studied, because very few species of this kingdom have been characterized. Ideally, more distant species should be studied, since Gallus gallus and Meleagris gallopavo, the only birds present in SelenoDB, belong to the same family, the Phaisanidae, and Colinus virginianus, even though proceeding from a different family, is within their same order (the Galliformes), so more phylogenetically distant species should be checked to follow the evolution of the selenogenome through the tree of life.