A general approach for activity-based protein profiling of oxidoreductases with redox-differentiated diarylhalonium warheads

Abstract

Activity-based protein profiling (ABPP) is a unique proteomic tool for measuring the activity of enzymes in their cellular context, which has been well established for enzyme classes exhibiting a characteristic nucleophilic residue (e.g., hydrolases). In contrast, the enzyme class of oxidoreductases has received less attention, as its members rely mainly on cofactors instead of nucleophilic amino acid residues for catalysis. ABPP probes have been designed for specific oxidoreductase subclasses, which rely on the oxidative conversion of the probes into strong electrophiles. Here we describe the development of ABPP probes for the simultaneous labeling of various subclasses of oxidoreductases. The probe warheads are based on hypervalent diarylhalonium salts, which show unique reactivity as their activation proceeds via a reductive mechanism resulting in aryl radicals leading to covalent labeling of liver proteins at several different amino acids in close proximity to the active sites. The redox potential of the probes can be tuned by isosteric replacement varying the halonium central atom. ABPP experiments with liver using 16 probes differing in warhead, linker, and structure revealed distinct overlapping profiles and broad substrate specificities of several probes. With their capability of multi oxidoreductase subclass labeling - including rare examples for the class of reductases - and their unique design, the herein reported probes offer new opportunities for the investigation of the "oxidoreductome" of microorganisms, plants, animal and human tissues.

Fig. 1. (A) Previous work based on the concept of “activation by oxidation” and selected examples thereof. (B) This work, based on the concept of “activation by reduction” and proposed, simplified mechanism for the mode of action via radical formation upon enzymatic reduction. Warheads are highlighted in orange and blue, the reporter tag click handle is highlighted in green.

Scheme 1. Synthesis of the warheads. Diphenyliodonium salts (A), diphenyleneiodonium salts (B) and diphenylbromonium and –chloronium salts (C).

Fig. 2. Crystallographic analysis of 1c (CCDC 2145617), 1f (CCDC 2149555) and 1h (CCDC 2145616).

Scheme 2. Synthesis of activity-based probes from diarylhalonium salts and azido-amine linkers. ^a Isolated yields after column chromatography. ^b Method A was used. ^c Method B was used. Warheads are highlighted in blue, reporter tag click-handles are highlighted in green.

Fig. 3. Cyclovoltammogram of 1c (blue curve, 30 mM concentration, solvent: MeCN, scan rate: 500 mV s⁻¹, supporting salt: Bu₄NClO₄ (0.1 M), working electrode: Pt). The red curves indicate the decrease of the reduction peak upon addition of NADH.

Scheme 3. Electron transfer reaction, spin trapping with PBN, and quenching with methyl t-butylacrylate (t-BAM).

Fig. 4. (A) ¹H-NMR spectrum of 1b (methanol-d₄/D₂O at 200 MHz). (B) ¹H-CIDNP spectrum recorded 30 min after mixing 1b and NADH. (C) ¹H-NMR recorded 30 min after the spectrum B. (D) CW-EPR during the reaction of 1b, NADH, and PBN overlaid with the simulated spectrum.

Fig. 5. (A) Fluorescent SDS-PAGE analysis of the labeling of various recombinant enzymes with probes 3aa and 3da (30 μM, 1 h; no = no probe added). Protein bands were visualized by fluorescent staining. (B) Total protein stains of activity-based gels. Proteins were visualized with Krypton fluorescent protein stain. ALDH2 = aldehyde dehydrogenase 2 (recombinant human), EUOX = eugenoloxidase (Rhodococcus jostii), CHO = choline oxidase (Arthrobacter nicotianae), FDH = formate dehydrogenase (Clostridium carboxidivorans), CYP 10v3/10v4 = cytochrome P450 10v3/10v4 (Phenylobacterium zucineum), CYP 13v3 = cytochrome P450 13v3 (linalool 8-monooxygenase; Mycobacterium intracellulare). CYP enzymes were used as crude cell lysates. Mark 12™ Unstained Standard (Thermo Fisher Scientific) was used as protein standard (=std).

Fig. 6. One-sided volcano plot for the labeling of murine liver with diphenyliodonium probe 3aa (30 μM, 1 h) showing the distribution of quantified proteins according to –log p value and log2-fold difference of proteins in probed (with probe 3aa) vs. non-probed samples (n = 6). Proteins above the line are considered statistically significant (adjusted p-value <0.01). Oxidoreductases are annotated.

Fig. 7. Hierarchical clustering and heat map, demonstrating the clusters with an abundance-scale (log2-fold enrichment). A total of 92 significant proteins (OXNAD1 to SLC25A20) are represented. Proteins present in higher amounts are shown in green, while those in lower amounts appear in red. The column tree shows the clustering of the probes 3aa–3ia (Scheme 2) based on their different profiles of enriched proteins.

Fig. 8. Side and zoomed views of the predicted AlphaFold structure of mCHDH (PDB Q8BJ64) depicting the potentially labeled amino acids (blue, shown in sticks) on the modified peptide fragment (yellow, first and last AA shown in sticks) and important active site residue (green, shown in sticks).