Sortase A substrate specificity in GBS pilus 2a cell wall anchoring

Abstract

Streptococcus agalactiae, also referred to as Group B Streptococcus (GBS), is one of the most common causes of life-threatening bacterial infections in infants. In recent years cell surface pili have been identified in several Gram-positive bacteria, including GBS, as important virulence factors and promising vaccine candidates. In GBS, three structurally distinct types of pili have been discovered (pilus 1, 2a and 2b), whose structural subunits are assembled in high-molecular weight polymers by specific class C sortases. In addition, the highly conserved housekeeping sortase A (SrtA), whose main role is to link surface proteins to bacterial cell wall peptidoglycan by a transpeptidation reaction, is also involved in pili cell wall anchoring in many bacteria. Through in vivo mutagenesis, we demonstrate that the LPXTG sorting signal of the minor ancillary protein (AP2) is essential for pilus 2a anchoring. We successfully produced a highly purified recombinant SrtA (SrtA(ΔN40)) able to specifically hydrolyze the sorting signal of pilus 2a minor ancillary protein (AP2-2a) and catalyze in vitro the transpeptidation reaction between peptidoglycan analogues and the LPXTG motif, using both synthetic fluorescent peptides and recombinant proteins. By contrast, SrtA(ΔN40) does not catalyze the transpeptidation reaction with substrate-peptides mimicking sorting signals of the other pilus 2a subunits (the backbone protein and the major ancillary protein). Thus, our results add further insight into the proposed model of GBS pilus 2a assembly, in which SrtA is required for pili cell wall covalent attachment, acting exclusively on the minor accessory pilin, representing the terminal subunit located at the base of the pilus.

Figure 1. Pilus proteins are mostly released into the supernatant in the AP2-2a_ΔLPKTG mutant strain.

Proteins were collected from FMC culture supernatants (B and D) or harvested cell pellets (A and C) of GBS strain 515 wild-type (wt), 515 knock-out strain for AP2-2a (ΔAP2) gene and ΔAP2 strains complemented with pAM-AP2_ΔLPKTG and with pAM-AP2 wild-type. Protein fractions were analysed by immunoblot stained with antibody specific for AP2-2a (A and B) and for the backbone protein BP-2a (C and D). Asterisks (*) indicate the monomeric form of AP2-2a and BP-2a. The equal quantity loaded in each well is verified by immunoblotting the same gel with a control antiserum that recognizes the protein PcsB of 47 kDa (indicated by a black arrow).

Figure 2. Molecular modeling of SrtA and assessment of metal ion binding by NMR spectroscopy.

Surface representation of S. agalactiae SrtA showing no similar anionic cleft (black circle) for calcium ion binding, as is the case for S. aureus SrtA. In red, negatively charged residues; in blue, positively charged residues (A). ¹H-¹⁵N-HSQC spectra recorded on ¹⁵N-SrtA_ΔN40 in 30 mM phosphate buffer, 1.5 mM DTT, pH 6.5 (B); 30 mM phosphate buffer, 1.5 mM DTT, 3 mM EDTA, pH 6.5 (C); 50 mM Hepes, pH 6.5 (D); 50 mM Hepes, pH 6.5, CaCl₂ at a protein∶calcium molar ratio of 1∶10 (E). Spectra were recorded at 600.13 MHz proton Larmor frequency and 298K. Both the addition of EDTA and calcium to the protein did not affect the chemical shift or intensity of the protein signals, indicating that no protein-calcium binding occurs.

Figure 3. FRET assay for in vitro enzymatic activity of SrtA_ΔN40.

(A) Hydrolytic activity of 3 µM SrtA_ΔN40 with 192 µM fluorogenic peptide Dabcyl-SFLPKTGM-Edans in HEPES buffer in which H₂O is the only nucleophile. (B) SrtA_ΔN40 (3 µM) catalyzes an in vitro reaction in the presence of 75 mM NH₂-Gly₃ and 256 µM of fluorogenic peptide Dabcyl-SFLPKTGM-Edans. The reaction rate is equivalent with or without 5 mM calcium ions.

Figure 4. SrtA_ΔN40 catalyzes in vitro transpeptidation reaction.

(A) RP-HPLC profile of the substrate peptide d-SFLPKTGM-e on a C4 column (the LPXTG motif is printed in bold). (B) The substrate peptide (20 µM) was incubated overnight with SrtA_ΔN40 (30 µM) in the presence of 5 mM NH₂-Gly₃ at RT. Then the reaction products were separated by RP-HPLC. The eluent was monitored by UV detection at 336 nm (red) and at 472 nm (black). (C) MALDI-TOF mass spectra of the reaction product confirmed the predicted molecular weight of the transpeptidation product d-SFLPKTGGG.

Figure 5. FRET assay with PI-2a peptides for substrate specificity analysis of SrtA_ΔN40.

The reaction solutions contained 256 µM of fluorescent peptide, 3 µM of SrtA_ΔN40 and 75 mM of triglycine as peptidoglycan analogue. The reactions were performed at 37°C in the assay buffer containing 20 mM HEPES, pH 7.5. We monitored fluorescence emission every 5 minutes and we observed an increase in fluorescence intensity only in the presence of AP2-2a peptide (filled squares), as opposed to BP-2a peptide (open squares) and AP1-2a peptide (filled triangles). In the presence of an additional variation of AP2-2a motif (AP2-2a_mut, open circles) the reaction did not occur.

Figure 6. Kinetic parameters of SrtA reaction with AP2-2a peptide.

(A) Progress curves of the transpeptidation reaction at various concentrations of AP2-2a peptide. The reaction solutions contained various concentrations of the peptide Dabcyl-SFLPKTGM-Edans (1, 2, 4, 8, 16, 32, 48, 64, 128, 256 and 384 µM), 75 mM triglycine and 3 µM SrtA_ΔN40. (B) Steady-state rate plotted against the substrate concentration generated a Km of 48.40 µM for the transpeptidation reaction.

Figure 7. RP-HPLC profile and MALDI-TOF mass spectra of sortase reaction with recombinant AP2-2a protein.

(A) Overnight incubation at room temperature of the reaction solution. (B) Starting point time of the reaction containing 30 µM of rAP2-2a, 30 µM of SrtA_ΔN40 and 5 mM of triglycine. Chromatographic tracing corresponds to UV absorption at 215 nm of a 100 µl aliquot assay. (C–D) Mass spectral analysis confirmed the presence of transpeptidation products after an overnight incubation at room temperature. The reaction solution contained 30 µM of rAP2-2a, 30 µM of SrtA_ΔN40 and 5 mM of triglycine. C upper panel shows the mass spectrum of the reaction product eluted at 27.84 min (shown in panel A), obtained by MALDI-TOF MS run in linear mode. In comparison, the uncleaved rAP2-2a eluted at 27.73 min (panel B) presented an m/z ratio of 28252.44 (C lower panel), in agreement with the expected average molecular mass of the recombinant protein with the lack of the initial methionine (28251.79 Da). Panel D shows the m/z signal of 1271.89 obtained in refectron mode from the reaction product eluted at 16.54 min (panel A), consistent with the theoretical monoisotopic mass of the peptide GMLEHHHHHH (1270.55 Da), generated by the rAP2-2a C-terminal cleavage after the transpeptidation reaction. Asterisks (*), oxidized form of the molecule (+16 Da). Triangles (∇), sodium adduct (+22 Da).

Figure 8. Sortase reactions with rBP-2a and rAP1-2a.

Reactions containing 30 µM of SrtA, 5 mM of triglycine and 30 µM of rBP-2a (panel A) or 30 µM of rAP1-2a (panel B) were analysed by RP-HPLC. Comparing time zero (red trace) and the overnight incubation at RT (black trace) of these reactions, we did not observe an extra peak due to the cleavage of PI-2a protein subunits by SrtA.