Figure 1.
β-Propellers are colored by the number of blades (4 = blue, 5 = light blue, 6 = green, 7 = orange, 8 = yellow, 10 = red). Most β-propellers are part of one connected cluster network and they are disconnected from most other clusters. A small number of β-propellers, primarily of viral origin, remain unconnected in sequence space, as discussed previously [15]. Clusters in dashed boxes were omitted in the detailed analysis after manual inspection (see “Spurious connections” in the cluster map section of the Methods). The purple groups are different superfamilies of the β-prism type I fold (b.77), unrelated to the β-prism type II fold discussed in this manuscript (see Fig. 2). The four clusters discussed in this manuscript are in red circles.
Figure 2.
Cluster map of β-propellers and their potential homologs.
Dots represent proteins, connections are similarities where darker means more similar. A) Potential β-propeller homologs are labeled and structural groups are colored as in Figure 1. The highly divergent WW domain YJQ8WW is annotated as well. B) The same cluster map as in A with β-propellers colored and labeled according to motifs and families; names too long to include in the figure are: * = Nitrous oxide reductase N-terminal domain, ** = prolyl oligopeptidase N-terminal domain, *** = Bacteriophage K1F endo-alpha-sialidase, and **** = glycoside hydrolase family 43, ***** = Hemagglutinin-Neuraminidase.
Figure 3.
A) Structure of the yeast IRE1-LD monomer with the five repeats detected by HHrepID colored and labeled. B) Sequence alignment of the 11-residue PQQ motif in blades 2–8 of BamB (blade 1 is a velcro blade and was omitted for not being continuous in this region) with the corresponding regions of yeast (y) and human (h) IRE1-LD repeats 1–5. At the top is shown the consensus PQQ motif [26]. Conserved motif positions have a gray background and residues adhering to the consensus are highlighted.
Figure 4.
Correlation between structure and sequence similarity in comparisons of the superfamilies analyzed in this study.
The comparisons were performed in all cases between a non-propeller fold and the closest β-propeller superfamily, as deduced from Figure 2. The panels show in orange: A) IRE1-LD vs. PQQ, B) β-prism type II vs. PQQ, C) β-pinwheels vs. WD40, and D) WW domains vs. PQQ. In each case, the comparison of the non-propeller fold to a background set of proteins consisting of the SCOP all-β class minus the superfamilies of this study is shown in blue as a reference (see also Methods). The plots represent the structure and sequence similarity of a pair of compared structures as a dot. The linear regression for each group of comparisons is shown as a dashed line, while the ellipses represent one, two, and three standard deviations around the mean.
Figure 5.
Analysis of type II β-prisms (BP2).
A) Structure of a BP2 (1XD5), colored blue to red from N- to C-terminus. B) Structural alignment of the three β-meanders of a BP2 (1XD5) and the eight blades of BamB (3Q7M) shown as a main chain trace. The 24 aligned residues result in an average RMSD of 1.28 Å. The side-chains of the conserved tryptophan residues in BamB and BP2 are located at the same position but with different orientations. C) Sequence alignment derived from the structural alignment in B. See Figure 3 for explanations.
Figure 6.
A) Structure of a closed-form β-pinwheel (1SUU). B) Topology diagrams of four consecutive β-strands in β-propellers and β-pinwheels. In β-propellers, the four β-strands form a single β-propeller blade. In β-pinwheels, β-strands B and C are part of the next blade, such that the four consecutive β-strands are part of two blades.
Figure 7.
A) Structure of a WW domain (1JMQ_A:13–42) bound to a proline-rich peptide (gray). The side-chains of the conserved tryptophan residues and the binding-site tyrosine are shown. B) Superimposition of the first WW domain of PRP40 (1O6W_A:1–29, rainbow coloring) with β-strands B–D of BamB β-propeller blade 2 (3Q7M_A:118–146, dark gray), shown as a main chain trace. The match is gapless and has an RMSD of 1.9 Å over 23 residues. C) Sequence alignment of the PQQ motif, BamB blades, and four WW domains. See Figure 3 for explanations.