Discovery of a new Pro-Pro endopeptidase, PPEP-2, provides mechanistic insights into the differences in substrate specificity within the PPEP family

Abstract

Pro-Pro endopeptidases (PPEPs) belong to a recently discovered family of proteases capable of hydrolyzing a Pro-Pro bond. The first member from the bacterial pathogen Clostridium difficile (PPEP-1) cleaves two C. difficile cell-surface proteins involved in adhesion, one of which is encoded by the gene adjacent to the ppep-1 gene. However, related PPEPs may exist in other bacteria and may shed light on substrate specificity in this enzyme family. Here, we report on the homolog of PPEP-1 in Paenibacillus alvei, which we denoted PPEP-2. We found that PPEP-2 is a secreted metalloprotease, which likewise cleaved a cell-surface protein encoded by an adjacent gene. However, the cleavage motif of PPEP-2, PLP↓PVP, is distinct from that of PPEP-1 (VNP↓PVP). As a result, an optimal substrate peptide for PPEP-2 was not cleaved by PPEP-1 and vice versa. To gain insight into the specificity mechanism of PPEP-2, we determined its crystal structure at 1.75 Å resolution and further confirmed the structure in solution using small-angle X-ray scattering (SAXS). We show that a four-amino-acid loop, which is distinct in PPEP-1 and -2 (GGST in PPEP-1 and SERV in PPEP-2), plays a crucial role in substrate specificity. A PPEP-2 variant, in which the four loop residues had been swapped for those from PPEP-1, displayed a shift in substrate specificity toward PPEP-1 substrates. Our results provide detailed insights into the PPEP-2 structure and the structural determinants of substrate specificity in this new family of PPEP proteases.

Keywords: PPEP; Paenibacillus alvei; Pro-Pro endopeptidase; bacterial adhesion; cell wall; metalloprotease; structural biology; substrate specificity; virulence factor.

Figure 1.

Genomic organization of PPEP-2 and its substrate (VMSP) in P. alvei DSM 29. A homolog of C. difficile PPEP-1 was identified in the genome of P. alvei (PPEP-2). Primary sequence alignment of PPEP-2 and PPEP-1 showed an overall sequence identity of 47% (upper part, N-terminal signal secretion sequences were removed for simplicity; numbering was according to PPEP-2). The predicted secondary structure is shown at the top of the alignment. α, α-helix; β, β-sheet; T, β-turns/coils; η, 3₁₀-helices). Next to the gene encoding PPEP-2, a gene was identified encoding a protein with a predicted von Willebrand factor A (VWFA) domain, several Muc-binding protein (MucBP) domains, and three surface layer homology (SLH) domains. We call this protein VMSP, according to these predicted domains (VWFA, Mucbp, Surface-layer homology Protein). Between the last MucBP domain and the first SLH domain, two PPEP-2 cleavage sites are found (PLPPVP). Arrowheads refer to PPEP-1 contacts (23).

Figure 2.

PPEP-2 is a Pro-Pro endopeptidase with a high specificity for the P2 and P3 positions of the cleavage site compared with PPEP-1. A, MALDI-ToF–MS spectrum of the cleavage products formed after a 1-h incubation of a synthetic VMSP-derived peptide (YPSSKPLPPVPPVQPLPPVPKLETS, amino acids 1098–1122) containing two PPEP-2–specific cleavage sites, with PPEP-2. B, comparison of substrate specificities of PPEP-2 (left panel) and PPEP-1 (right panel). The progress curves show the increase of fluorescence during a 1-h incubation upon the protease-mediated cleavages (50 μm FRET substrate peptide and 200 ng of enzyme). Black, cleavage of a FRET peptide containing PLPPVP; red, cleavage of a FRET peptide containing VNPPVP.

Figure 3.

Atomic structure of PPEP-2. Atomic structure of chain A is in cartoon representation. Definition of the domains is the same as for PPEP-1: N-terminal domain (NTD), blue; active site (helix α4), yellow; C-terminal domain (CTD), green. Zinc-coordinating residues and residues involved in catalysis are shown as sticks. Zinc ion is shown as a sphere. For more details on the crystal unit, see Table 1 and the supporting information.

Figure 4.

Analysis of PPEP-1 and PPEP-2 using SAXS. A, overlay of the SAXS profiles obtained for PPEP-1 and PPEP-2. B, fit from the best-fitting chain B of the PPEP-2 crystal structure to the PPEP-2 experimental curve. C, superposition of the PPEP-2 crystal structure chain B and the ab initio envelope obtained using DAMMIN (NSD = 1.45) (29).

Figure 5.

Structural comparison of substrate recognition in PPEP-1 and PPEP-2. A, superposition of the substrate-bound PPEP-1 structure (PDB code 5A0X, gray) with PPEP-2 (this work, chain B, cyan). The substrate peptide bound to PPEP-1 is shown in orange. The β3/β4 loop is shown in yellow for PPEP-1 and in red for PPEP-2. The S-loop is shown in green for PPEP-1 and in blue for PPEP-2. Hydrogen bonds between the peptide and PPEP-1 are shown as magenta dashed lines. B, close-up of the substrate peptide bound to PPEP-2 (yellow, model). For clarity, only the nonprime moiety of the modeled peptide is shown. A proline at the P3 position, as present in the optimal PPEP-2 substrate peptide, produces an additional kink in the polypeptide. As a result, the upstream polypeptide chain deviates away from the salt bridge formed by residues Glu-113 and Arg-145. For comparison, the substrate peptide bound to PPEP-1 (gray, crystal structure), with a Val at the P3 position, is overlaid. With PPEP-2, this conformation results in a sterical clash with the salt bridge.