plosone-phylo
pone.0002616.g003.png
Phylogenetic trees of AFPs and related lectins as well as selected teleost fishes.(A) Bayesian 50% majority rule consensus tree (from 7500 sampled trees) based on the protein sequences aligned in Figure 2. The alignment was truncated to the beginning of the first helix (α0) to the end of the sea raven sequence. Branch lengths represent the expected fraction of changes, and Bayesian posterior probabilities above 50% and selected percent identities (italics) are indicated near the nodes. The highest identity between the AFPs and any other C-type lectin domain is 38%. Herring AFP is 84% and 85% identical to Japanese and rainbow smelt AFPs, respectively. The two smelt sequences share 79% identity. (b) Phylogenetic tree of teleosts based on ribosomal 16S RNA sequence comparison. Species are referred to by their common names. Those that are known to produce AFPs have the type of AFP they produce indicated alongside (AFGP, antifreeze glycoprotein; I, alanine-rich alpha-helical; II, lectin-like; III, beta-clip fold; IV, helix bundle). Those that produce type II AFP are highlighted. The rainbow smelt and herring sequences were double-checked by amplification from the DNAs used in this study and were almost identical to the sequences reported in the database. The non-teleost fish, the bowfin, was used to root the tree and only posterior probabilities >50% are shown. Putative type II gene loses (presuming the presence of the gene in the common ancestor) are indicated by grey stars. A further gene loss on the Takifugu/masked triggerfish lineage is not postulated since these fish group with yellow perch and Antarctic toothfish according to the phylogeny of Miya et al. [22].
pone.0002616.g003.png
Phylogenetic trees of AFPs and related lectins as well as selected teleost fishes.(A) Bayesian 50% majority rule consensus tree (from 7500 sampled trees) based on the protein sequences aligned in Figure 2. The alignment was truncated to the beginning of the first helix (α0) to the end of the sea raven sequence. Branch lengths represent the expected fraction of changes, and Bayesian posterior probabilities above 50% and selected percent identities (italics) are indicated near the nodes. The highest identity between the AFPs and any other C-type lectin domain is 38%. Herring AFP is 84% and 85% identical to Japanese and rainbow smelt AFPs, respectively. The two smelt sequences share 79% identity. (b) Phylogenetic tree of teleosts based on ribosomal 16S RNA sequence comparison. Species are referred to by their common names. Those that are known to produce AFPs have the type of AFP they produce indicated alongside (AFGP, antifreeze glycoprotein; I, alanine-rich alpha-helical; II, lectin-like; III, beta-clip fold; IV, helix bundle). Those that produce type II AFP are highlighted. The rainbow smelt and herring sequences were double-checked by amplification from the DNAs used in this study and were almost identical to the sequences reported in the database. The non-teleost fish, the bowfin, was used to root the tree and only posterior probabilities >50% are shown. Putative type II gene loses (presuming the presence of the gene in the common ancestor) are indicated by grey stars. A further gene loss on the Takifugu/masked triggerfish lineage is not postulated since these fish group with yellow perch and Antarctic toothfish according to the phylogeny of Miya et al. [22].