Figure 3: General features of sumoylation at the residue level.

a) Distribution of sites between sumoylated proteins.(b) Amino-acid context of sumoylated lysines. Four pLogo graphs display the amino-acid sequence context of all sumoylated lysines; of KxE/KxD consensus-site lysines; of lysines within the ExK/DxK inverted consensus site and of sites not conforming to either consensus (residual sites). The numbers in parentheses above each graph indicate the total number of sites in the category (first number) and, among them, the number of sites that covered the full 13-amino acids window required for the pLogo analysis (second number). The log-odds of the binomial probability (π) are represented on the y-axis. Significance threshold values of 3.68 (P<0.05) are shown in red. (c) Sumoylated residues identified in this study are shown in yellow in a cartoon structure of S. cerevisiae topoisomerase 2 (PDB#4GFH; 94% identical to S. pombe). The portion of the protein whose structure was solved is coloured in the S. cerevisiae Top2 representation on the right. Enlarged cartoon views show the tendency of sumoylation sites to fall into coils or end of helices, the proximity of sumoylation sites in three-dimension and the proximity of sites to DNA. (d) Distribution of 82 sumoylated lysines identified in this study and mapped onto existing crystal structures using PyMOL. The structures are shown in Supplementary Figs 4 and 5 and the sites are listed in Supplementary Data 7. ‘End-of-helix’ and ‘end-of-sheet’ refer to lysines at three or less amino acids from the end of a helix or sheet. (e) Distribution of lysines in secondary structures predicted with PSIPRED⁷⁰ for respective sumoylated lysines, and for all the lysines in proteins with a sumoylated lysine. *P<0.001. P values were calculated with a hypergeometric probability test. See also Supplementary Data 8.