Fig. 2.

Kinetic subclassification of cleavage site specificity and structural accessibility. A, Subclustering of GluC-generated neo-N termini. Peptides assigned to cluster N.1 with a membership value of ≥0.8 were assigned to three subclusters with slow, medium and fast increase in abundance on GluC incubation. Colorkey indicates membership value α. ctrl: 12 h and 16 h controls. B, IceLogo analysis of GluC cleavage sites corresponding to peptides assigned to each subcluster. Less efficient cleavages C-terminal to aspartate were assigned to subcluster N.1.1 comprising slow cleavage events, whereas medium and fast clusters comprised neo-N-terminal peptides derived from efficient cleavages after glutamate. C, Secondary structure preferences for GluC cleavages in each subcluster. Secondary structure elements around the scissile bond were predicted by PROTEUS2 and aligned for cleavage sites in each subcluster. For all clusters only cleavages C-terminal to glutamate were included. High cleavage efficiency in subcluster N.1.3 is correlated with prevalence of extended loops. L: loop; A: helix; B: sheet. D, Correlation of structural elements and cleavage efficiency within the same protein. Models from the SWISS-MODEL Repository for P63260 (left) and P60335 (right) based on templates 3ub5A (P63260), 2jzx (P60335; residues 11–169) and 1wvn (P60335; residues 278–348), respectively. GluC cleavage sites identified by 8plex-iTRAQ-TAILS are indicated (P1 (E): red; P1′: yellow). Cleavages in loops (PIYE¹⁶⁷.¹⁶⁸GYAL and VMLE¹⁶⁸.¹⁶⁹TLSQ) were more efficient than in a helix (EYDE³⁶⁴.³⁶⁵SGPS) or a beta sheet (TTHE²⁸³.²⁸⁴LTIP), indicating discrimination between secondary structure elements by time-resolved iTRAQ-TAILS analysis.