Figure 4

Multiple sequence alignment of FS domains from proteins across different species. Based on the FS domain structures of the two monomers, the secondary structures are indicated with arrows (β-strands) and coils (α-helices) above the appropriate sequences. Identical residues are highlighted in red. Disulfide bond forming cysteines are indicated with green numbers. Disulfide bond substitution, potential hydrogen bond forming residues (T190 and Q330) in amphiF-spondin are colored in blue. The integrin-binding LEV triplet found in human, mouse and rat mindins are highlighted in magenta. The sequences used in the alignment include human F-spondin (gi:110347423), rat F-spondin (gi:544353), chicken F-spondin (gi:45382337), Xenopus F-spondin (gi:544354), zebrafish F-spondin-1 (gi:18859413), zebrafish F-spondin-2 (gi:2529227), amphiF-spondin (gi:3319874), Drosophila melanogaster F-spondin (gi: 19922086), human mindin (gi: 52783469), mouse mindin (gi: 52783454, rat mindin (gi: 52783424) and zebrafish mindin-1 and -2 (gi: 82069286 and gi: 82069288). The alignment was generated with the program MultAlin[38] and ESPript[39]. In Drosophila melanogaster F-spondin (D. M-spondin), the large, likely unstructured, Ser, Gly, Thr and Ala-rich insertion (VQMQLQSQMQAGKSPSGGISSGTTSFNATAASTATPTGGSGGSGGSGGSGGTGTTTAE) between α3 and β2 is not shown in the alignment.