Additional file 2: of Diversity of opisthokont septin proteins reveals structural constraints and conserved motifs

Figure S1. Maximum likelihood phylogenetic analysis with RAxML software. Node values represent bootstrap support. Protein names are given for septins supported by experimental evidence. Aspergillus and Drosophila sequences used to recognize septin groups are in bold. Coiled-coil domain predictions, black representing p < 0.05 and grey p < 0.10, found to the right of names. Domain predictions for proteins longer than 600 residues have been shortened with diagonal lines. Figure S2. Bayesian phylogenetic analysis. Node values represent posterior probabilities. Protein names are shown for those septins with experimental evidence. Aspergillus and Drosophila sequences used to recognize septin groups are in bold. Coiled-coil domain prediction, black representing p < 0.05 and grey p < 0.10, shown to the right of names. Coiled-coil predictions for proteins longer than 600 residues have been shortened with diagonal lines. Figure S3. Bayesian phylogeny with jPRIME software. Topology represents maximum clade credibility tree; node values represent bootstrap support. Protein names are shown for those septins with experimental evidence. Aspergillus and Drosophila sequences used to recognize septin groups are in bold. Coiled-coil domain prediction, black representing p < 0.05 and grey p < 0.10 to the right of septin names. Coiled-coil predictions for proteins longer than 600 residues have been shortened with diagonal lines. Figure S4. Ancestral state reconstructions for presence of septin groups inferred using Mesquite with the MK1 symmetrical model. Shading of pie charts at nodes represent proportional likelihood of a node containing a member of that septin group. Statistical test showing that MK1 could not be rejected appears below state reconstructions. This test supports assuming a single rate of change for gains and losses. Figure S5. Interacting residues. A) Interacting residues found based on modelling the 5 individual crystal structures. Red or blue shading indicates the proportion of taxa for which a given residue interacts in the NC or G interface, respectively. Note that no single crystal structure alone can be used to assess all interface regions due to low resolution portions in each crystal. Crystal structures used were as follows: 3TW4 (Human Sept7) provided the Group 1–1 Homodimer interface. 2QA5 (Human Sept 2/6/7 hexamer) provided the Group 2–2 homodimer G and NC interface, and the Group 2–1 NC interface. 2QAG (Human Sept2 dimer) provided the Group 2–2 homodimer G interface. B) Interface conservation as in Fig. 3A, with the solid line representing Shannon-Jensen sequence conservation; shading indicates values above 0.5. GTP-biding residues are indicated with black arrows, with sequential residues having overlapping arrows. The red and blue columns indicate the highest value for a position from the individual crystal structures in A). Figure S6. COBALT alignment of representative septins from the 7 groups, showing location of conserved regions and interface regions in a representative septin from each of the seven groups. Green cylinders represent position of alpha helices, and pink arrow indicate beta sheets. Consistently interacting residues are indicated by blue for G interface, and red for NC interface. Motifs identified by Pan et al., [15] are outlined in black boxes. (ZIP 2531 kb)