Superoxide
dismutase
Trypanosomatids have at least three
superoxide dismutase (SOD) activities located in the cytosol, the
mitochondrion and the glycosomes, respectively (Opperdoes
et al., 1977). So far, the sequences
for only two genes, A and B, both of the Fe-type have been described
for Trypanosoma and Leishmania spp. The A and B genes
encode distant but homologous proteins, which have probably entered
the trypanosomatid ancestor with an early event of horizontal
transfer.
Further details of the analysis
A Trypanosoma cruzi SOD was compared with all the sequences of the SwissProt database indexed at European Bioinformatics Institute (EBI, Hinxton UK), using the NCBI BLASTP program and the BLOSUM 62 matrix (http://www2.ebi.ac.uk/blastall/) . Click here to inspect the BLASTP output file. The best scoring sequences were those from two cyanobacterial isolates of Synechocystis and Synechococcus which had 45 and 46 % identity with the trypanosome sequence.
From the best scoring sequences (E values of e-10 or better) containing only bacterial Fe and Mn SODs plus a few plant chloroplast sequences and that of E. histolytica, incomplete sequences were removed. Remaining sequences were selected for alignment in Clustal and several trypanosomatid sequences were added by the profile alignment utility of Clustal.
The resulting alignment was freed of gaps and incomplete or duplicate sequences were removed resulting in an alignment of 41 sequences with 118 sites. The program Clustal was used to calculate a table with uncorrected pairwise distances. Pairwise percentages of identity of the trypanosomatid sequences and the other sequences after removal of gaps ranged from 50 to 31.
The alignment in Phylip format containing 41 sequences and 118 sites was used for likelihood mapping as implemented in PUZZLE version 4.0.1. The result indicated that due to the limited number of sites (118) the dataset contained a weak phylogenetic signal with as much as 9.2% of all quartets unresolved and with only 84 % of fully resolved quartets. Four-cluster likelihood mapping showed that only 18% of all quartets supported a branching of the trypanosomatids (a) together with the chloroplast and cyanobacterial SODs (b), while 33% of quartets supported a branching with the manganese SODs (d) and only 11% with the FeSODs.
A bootstrapped neigbor-joining tree was created in Clustal using the alignment in Phylip format above. The NJ tree (10 000 bootstrap samplings) showed a branching of the Trypanosomatidae close to the chloroplasts and the cyanobacterial SODs, but the tree was not robust and gave, except for the separation between the two clades of Mn- and Fe-SODs, no bootstrap support for any of the deep branching taxa, including the trypanosomatid, chloroplast and cyanobacterial taxa. All trypanosomatid sequences clustered together but again with very low bootstrap support (46%).
A maximum likelihood tree (10 000 puzzle steps) gave a star-like tree with no differentiation in the branching order of most of the groups of taxa. Also the maximal parsimony tree (100 bootstrap samplings) gave a tree with very low bootstrap support for the deeply branching taxa.
In conclusion, there is strong support for an event of horizontal transfer of a prokaryotic SOD from a bacterium to a trypanosomatid ancestor. All trypanosomatid SODs cluster with the Fe- and Mn-SODs of bacteria and the only other eukaryotic SODs found in this group (apart from that of Entamoeba histolytica, which must also represent an event of horizontal gene transfer) are those of cyanobacterial origin now found in plant chloroplasts. However, despite the fact that a BlastP search gave the best scores to the cyanobacterial sequences, the dataset contained insufficient phylogenetic signal to be able to identify the actual prokaryotic ancestor for the trypanosomatid SOD. The low quality of the dataset is also illustrated by the fact that bootstrap support for monophyly of the trypanosomatid SODs in the NJ and MP trees was below 50%.