Mass spectrometry-based determination of Kallikrein-related peptidase 7 (KLK7) cleavage preferences and subsite dependency

Abstract

The cleavage preferences of Kallikrein-related peptidase 7 (KLK7) have previously been delineated using synthetic peptide libraries of fixed length, or single protein chains and have suggested that KLK7 exerts a chymotryptic-like cleavage preference. Due to the short length of the peptides utilised, only a limited number of subsites have however been assessed. To determine the subsite preferences of KLK7 in a global setting, we used a mass spectrometry (MS)-based in-depth proteomics approach that utilises human proteome-derived peptide libraries of varying length, termed Proteomic Identification of protease Cleavage Sites (PICS). Consistent with previous findings, KLK7 was found to exert chymotryptic-like cleavage preferences. KLK7 subsite preferences were also characterised in the P2-P2' region, demonstrating a preference for hydrophobic residues in the non-prime and hydrophilic residues in the prime subsites. Interestingly, single catalytic triad mutant KLK7 (mKLK7; S195A) also showed residual catalytic activity (k_cat/K_M = 7.93 × 10² s^-1M^-1). Catalytic inactivity of KLK7 was however achieved by additional mutation in this region (D102N). In addition to characterising the cleavage preferences of KLK7, our data thereby also suggests that the use of double catalytic triad mutants should be employed as more appropriate negative controls in future investigations of KLK7, especially when highly sensitive MS-based approaches are employed.

Figure 1

KLK7 and mKLK7 protein identity. (a) Surface exposed residues corresponding to the catalytic triad (His57, Asp102 and Ser195; Schechter and Berger notation) of KLK7 were identified on the three dimensional (3D) structure available in Protein Data Bank (PDB, accession: 2QXG.pdb in standard serine protease orientation) using SPDBV v4.10. (b) In addition, S195A mutation was made using the mutation tool on SPDBV v4.10. Alongside, are the peptides identified in the MS analysis (Underlined; identified with a 99% protein probability and 95% peptide probability cut off) of trypsin-digested (a) active KLK7 and (b) mKLK7 aligned with the KLK7 full length protein (UniProtKB; P49862-1). The pro-signal peptide (in blue box) was not included in the expression construct, thus obtaining mature amino acid sequences for these peptidases. Expected mutation, S195A was identified in the MS-identified mKLK7 peptide sequence as highlighted in red bold letter.

Figure 2

Activity of KLK7 and mKLK7. (a) Coomassie-stained 1% casein zymography: cleared area at approximately 25 kDa shows digestion of casein by active KLK7 while mKLK7 has not digested casein, suggesting no activity. (b) Activity with a KLK7-specific peptide substrate MeO-Suc-Arg-Pro-Tyr-MCA: time versus absorbance (corrected mean ± SD) plot showing activity of KLK7 compared to no activity with mKLK7 and substrate only control (n = 6; mean ± SD). (c) Activity against a KLK7 protein substrate fibronectin: KLK7 and mKLK7 incubated with fibronectin (KLK7/FN: 1/10 molar ratio) visualised by silver-stained SDS-PAGE. Black arrow heads indicate KLK7-generated cleavage fragments. No cleavage was observed for mKLK7. (d) For KLK7 and (f) mKLK7 the percentage amino acid occurrences in P6-P6′ derived from both libraries were calculated and are shown in the form of a two-dimensional heat map. Both libraries confirm the predominant KLK7 specificity for Y, L and F in P1. The mKLK7-treated (f) tryptic and GluC libraries confirm the mKLK7 specificity for Y in P1. White line depicts the scissile peptide bond between P1 and P1′. (e) KLK7 and (g) mKLK7 cleavage sites derived from both libraries are displayed as icelogos by analysing their multiple sequence alignments. The height of the single amino acid residue (in bits) reflects its occurrence rate for each position in P10-P10′. Amino acids are coloured according to their physico-chemical properties.

Figure 3

Activity profile of the double mutant KLK7 (dmKLK7). (a) Surface exposed residues corresponding to the catalytic triad of dmKLK7 (His57, Asn102 and Ala195) are shown on the 3D-structure available in PDB (accession: 2QXG.pdb in standard serine protease orientation) using SPDBV v4.10. Alongside, are the peptides identified in the MS-analysis (underlined) aligned with the KLK7 full length protein (UniProtKB; P49862-1). S195A and D102N mutations are as highlighted in red bold letters. (b) dmKLK7 cleavage site preferences in the tryptic-PICS library (1:50; enzyme/library) in the form of a 2D-heat map showing scattered cleavage specificities from P6-P6′ with no chymotryptic-like cleavage site specificities. (c) Casein-zymogram comparing active KLK7, mKLK7 and dmKLK7; (d) MeO-Suc-Arg-Pro-Tyr-MCA peptide substrate assay (n = 6, mean ± SD) with mKLK7 and dmKLK7; and (e) silver-stained SDS-PAGE with the protein substrate, FN, comparing active KLK (1/10 to 1/1000) with dmKLK7 .

Figure 4

KLK family sequence similarity. (a) Human KLK full length protein sequences (UniProtKB) were aligned using Clustal Omega. Phylogram indicated high sequence similarity between KLK7, KLK4 and KLK5 (in box). (b) KLK4, 5 and 7 sequence alignment represents high sequence similarity adjacent to the catalytic triad residues (His57, Asp102 and Ser195 in red). N-termini of mature KLKs starting at Ile16 is shown in bold letters. “*”, a single, fully conserved residue; “:”, conservation between groups of strongly similar properties-scoring >0.5^; “ . ”, conservation between groups of weakly similar properties-scoring ≤0.5^; ^Gonnet PAM250 matrix. (c) mKLK4 showed residual activity against d-VLR-AFC peptide substrate compared to the dmKLK4.