Alignment of partial SH2 domains of Stat proteins with homologous SH2 domains of known structure. All sequences were downloaded from GenBank, except for SH2 domain of Stat3, which was used in the search query. The numbering of the peptide fragments and accession codes are as follows: Grb2 67-140, P29354; Syp 120-202,P35235; p56 Lck 134-213, P06239; v-Src 154-234, P15054; Stat3 590-661; Stat1, 572-646, P42224; Stat2, 387-459, P52630; Stat4, 579-651, P42228; Stat5A 599-665, P42229; Stat5B 599-665, P51692; and Stat6 543-611,P42226. STAT sequences (in black) were aligned using CLUSTAL, as were nonSTAT sequences (in gray). The relative alignments between STAT and non-STAT proteins were based on alignments of Chen et al,58which used direct comparison of the crystal structures of the SH2 domains of v-Src and Stat1. Secondary structural elements observed in structural studies of v-Src and Stat1 SH2 domains are indicated below the sequences (in gray and black, respectively); β-sheets are indicated as arrows, and -helices as boxes with diagonal stripes. Residue sidechains, which have been shown in high resolution studies to interact directly with the phosphotyrosine of the ligand, are boxed and highlighted. The v-src sidechains of which form the binding pocket for the +3 residue of the ligand are highlighted (Y202 and I214). Residues of Stat3, which are hypothesized to interact with bound ligand, are also highlighted (E638, Y640, and Y657).