Figure 1.
Comparison of the structure of GSTE clusters in the three Anopheles species.
Transcriptional orientation of each GST gene is shown by an arrow. The size of each gene is indicated in the boxes and the intergenic region size is shown above the lines. A) Anopheles stephensi; B) Anopheles funestus; C) Anopheles plumbeus and D) Anopheles gambiae. * indicates that gene sequence is not complete. Arrows above the genes indicate orthologs with Aedes aegypti. GSTE8 and GSTE3 were not amplified in An stephensi, An. funestus and An. plumbeus.
Figure 2.
Alignment of amino acid residues of the GST epsilon class in Anopheles species.
Residue numbering for each sequence is shown at the top. Conserved residues are shaded (>80%). The conserved region in the C-terminal domain is boxed. * represents amino acid highly conserved among GSTs. Sites under selection have been highlighted in colour. The three shorter sequences excluded from the second analysis are in red type. The highlighted sites were inferred by the Bayes Emperical Bayes method to have ω>1. The probability of the site being assigned to a class with ω>1 is indicated by the color of the shading: yellow: P>0.99 in both GSTall and GSTnoe6pfd; red: 0.95<P<0.99 in both GSTall and GSTnoe6pfd; green: 0.95<P<0.99 in GSTnoe6pfd only; blue: 0.95<P<0.99 in GSTall only; pink: 0.95<P<0.99 in GSTall, P>0.99 in GSTnoe6pfd; grey: P>0.99 in GSTall, 0.95<P<0.99 in GSTnoe6pfd.
Figure 3.
Phylogenetic relationship of Anopheles epsilon class GSTs.
Maximum likelihood trees for epsilon class GSTs used in PAML analysis. Branch support is given as a percentage of 500 bootstrap replicates. A) For all available sequences and B) excluding truncated sequences for GSTE6 for Anopheles funestus, An. plumbeus and An. darlingi. The foreground branch used in the branch and branch-site models is marked #1. Note that while midpoint rooted trees are shown here for ease of reading; unrooted trees were used in PAML analysis.
Table 1.
Variation in intron size and amino acid (AA) sequences for An. gambiae, An. funestus, An. stephensi and An. plumbeus.
Figure 4.
Sequence differences between ApGSTE2 and ApGSTE2B mapped onto structural models of each.
A) Overall distribution of differences. All differences are shown with respect to the side chains present in ApGSTE2 on its structural model, represented as both cartoon and surface. Ball and stick representation is used for ligands (white carbon for DDT, as modelled by Wang et al., 2008 [29] into AgGSTE2, magenta for GSH present in crystal structures of AgGSTE2). The position of the second chain of the dimer is shown as a cyan cartoon. B) Cross-eyed stereo close-up of the catalytic site showing nearby sequence differences as sticks (green for ApGSTE2, purple for ApGSTE2B) and ligands as in A). Sequence differences are labelled, for example, as Q41H to indicate that Gln41 in ApGSTE2 is replaced by His in ApGSTE2B.
Table 2.
Likelihood ratio test of positive selection at sites in the GSTE cluster.
Figure 5.
Sites inferred to be under positive selection in GSTE5.
Sites under positive selection: ω>1 with P>0.95 in both GSTall and GST no e6 pfd datasets; (see Figure 2) are shown as sticks on a structural model of AgGSTE5. Ball and stick representation is used for ligands (white carbon for DDT, magenta for GSH – see Wang et al. [29]. Phe212, at the heart of the H site is dark blue, other positions cyan.
Table 3.
Location of GSTE polyadenylation signals and 3′ UTR lengths in An. stephensi, An. funestus and An. plumbeus.