Protein homeostasis in bacteria is regulated by proteases such as the tetradecameric caseinolytic protease P (ClpP). Although substrates of ClpP have been successfully deciphered in genetically engineered cells, methods which directly trap processed proteins within native cells are elusive. Here we introduce an in situ trapping strategy which utilizes trifunctional probes that bind to the active site serine of ClpP and capture adjacent substrates with an attached photocrosslinking moiety. Substrate deconvolution by mass spectrometry (MS) is accomplished after enrichment using an alkyne handle. We show that our two traps bind substoichiometrically to ClpP, retain protease activity, exhibit unprecedented selectivity for Staphylococcus aureus ClpP in living cells and capture numerous known and novel substrates. The exemplary validation of trapped hits using a targeted proteomics approach confirmed the fidelity of this technology. In conclusion, we provide a novel chemical platform suited to discover serine protease substrates beyond genetic engineering.

Accurate quantification of 8 selected identified ClpP substrates in S. aureus WT vs. S aureus S98A (proteolytically inactive).

Sample preparation 120 μL adjusted lysate of S. aureus NCTC8325 wt and ClpP S98A treated with chloramphenicol at OD600 = 1.0 was taken from the experiment described above. Upon EtOH precipitation on carboxylated beads, the lysate was cleaned up using the SP3 method (as described above) on an automated handling system (Microlab Prep, Hamilton). Samples were desalted using Stage tips (styrenedivinylbenzene reversed phase sulfonate filter, 2 layers, 47 mm, Empore) according to a published protocol.[14] Samples were lyophilized and the remaining peptides solved in 125 μL 1% (v/v) aqueous FA. For internal retention time reference, iRT peptides (Biognosys) were added 1:20 to each sample. Data acquisition on Orbitrap Fusion For the quantification of the 8 selected proteins in S. aureus ClpP S98A vs. S. aureus wt, targeted PRM measurements were performed on 3-6 selected peptides per protein (see isolation list below). For that an UltiMate 3000 nano HPLC system (Dionex) equipped with an Acclaim C18 PepMap100 (75 μm ID x 2 cm) trap column and 25 cm Aurora Series emitter column (25 cm × 75 μm ID, 1.6 μm FSC C18) (Ionopticks, column oven heated to 40 °C) in an EASY-spray setting coupled to an Orbitrap Fusion (Thermo Fisher Scientific Inc.) was used. 1 μL peptide sample was loaded on the trap and washed with 0.1% (v/v) TFA, then transferred to the analytical column (buffer A: H2O with 0.1% (v/v) FA, buffer B: MeCN with 0.1% (v/v) FA, flow 0.4 μL/min, gradient: to 5% buffer B in 7 min, from 5% to 22% buffer B in 105 min, then to 32% buffer B in 10 min, to 90% buffer B in 10 min and hold at 90% buffer B for 10 min, then to 5% buffer B in 0.1 min and hold 5% buffer B for 9.9 min) and ionized by nanospray ionization (NSI) with a spray voltage of 1.8 kV and a capillary temperature of 275 °C. The Orbitrap Fusion was operated in parallel reaction monitoring mode. Full scan MS1 spectra were recorded in the 300 to 1500 m/z range with a resolution of 60,000, using an AGC target set to 2e5 and maximum injection time of 25 ms. Targeted MS2 spectra were acquired in the 100 to 2000 m/z range with a resolution of 30,000, an AGC target value of 1.25e4, maximum injection time of 20 ms, and an HCD collision energy of 30%. The RT windows used for the scheduled PRM experiments was set to 5 min. Estimated maximal cycle time was 576 ms.