File S1 - Early Chordate Origin of the Vertebrate Integrin αI Domains

Table S1: Sequences utilized in the phylogenetic analysis. Table S2. Residues in the α1I domain structure within 4.2 Å (non-hydrogen atoms) of the bound GLOGEN tripeptide (NMR structure; [21]) and equivalent residues in the human αI domains and the sequence fragments from the lamprey and hagfish. Where available, the sequence numbering is from a three-dimensional structure (PDB codes and resolution are indicated for the known X-ray structures). The metal ion at MIDAS is covalently bound to the tripeptide ligand. Residues from MIDAS (S13, S15, T81 and D114 in α1I, 3M32) are in italics and one residue, D11 in α1I (not listed) is absolutely conserved across all of the sequences. In the X-ray structure of α1I (PDB code: 1PT6; [79]) and this residue (D150 in 1PT6) binds to the metal at MIDAS via an intervening water molecule (WAT603). Table S3. Residues in the αLI domain structure within 4.2 Å (non-hydrogen atoms) of the bound ICAM and equivalent residues in the human αI domains and the sequence fragments from the lamprey and hagfish. Where available, the sequence numbering is from a three-dimensional structure (PDB codes and resolution are indicated for the known X-ray structures). The metal ion at MIDAS is covalently bound to the tripeptide ligand. Residues from MIDAS (S139, S141 and T206 in αLI, 1T0P) are in italics and two residues, D137 and D239 in αLI (not listed), are conserved across all of the sequences and functions to bind the metal at MIDAS via a water molecule (WAT943). Figure S1. Phylogenetic analysis of integrin sequences with the Bayesian method using MrBayes and based on the species and sequences listed in Tables 1 and S1. (A) Full-length sequence alignment of integrin α subunits his dataset contains the nearly full-length integrin α subunit from the sea lamprey Pma_f3 (highlighted in bold). (B) Tree based on the aligned common sequence region in all three lamprey sequence fragments Pma_f1, Pma_f2 and Pma_f3 (highlighted in bold). (C) Tree based on the alignment of the integrin αI domain sequences; this dataset includes the three lamprey αI domain sequences Pma_f1, Pma_f2 and Pma_f3 (highlighted in bold) and the hagfish fragment Ebu_f (highlighted in bold). Bayesian phylogenetic trees were constructed by implementing the Whelan and Goldman substitution matrix with frequency model and gamma distribution with invariant sites (WAG+I+G+F). Statistical support, in the form of the percentage posterior probability, was obtained with a MCMC run of 106 generations and the resulting percentage support value is indicated at each node. Figure S2. Phylogenetic analysis of integrin sequences with the Neighbor joining method using MEGA and based on the species and sequences listed in Tables 1 and S1. (A) Full-length sequence alignment of integrin α subunits his dataset contains the nearly full-length integrin α subunit from the sea lamprey Pma_f3 (highlighted in bold). (B) Tree based on the aligned common sequence region in all three lamprey sequence fragments Pma_f1, Pma_f2 and Pma_f3 (highlighted in bold). (C) Tree based on the alignment of the integrin αI domain sequences; this dataset includes the three lamprey αI domain sequences Pma_f1, Pma_f2 and Pma_f3 (highlighted in bold) and the hagfish fragment Ebu_f (highlighted in bold). Neighbor joining trees were constructed by implementing the Jones and Thornton (JTT) matrix. Statistical support for each phylogenetic tree was obtained with 1000 bootstrap replicates and the percentage bootstrap support value is indicated at each node. Figure S3. SDS PAGE of Pma_f1-3, human wild-type α2I, GST and molecular weight standards (st). SDS PAGE was run according to manufacturer's instructions using the GE Healthcare PhastSystem (GE, USA) and 8-25% gradient gel. Protein samples were adjusted to 300 ng/ml and the sample size was 1 µl. The gel was stained with Coomassie Brilliant Blue. (DOC)