Comparative analysis of cleavable azobenzene-based affinity tags for bioorthogonal chemical proteomics

Abstract

The advances in bioorthogonal ligation methods have provided new opportunities for proteomic analysis of newly synthesized proteins, posttranslational modifications, and specific enzyme families using azide/alkyne-functionalized chemical reporters and activity-based probes. Efficient enrichment and elution of azide/alkyne-labeled proteins with selectively cleavable affinity tags are essential for protein identification and quantification applications. Here, we report the synthesis and comparative analysis of Na₂S₂O₄-cleavable azobenzene-based affinity tags for bioorthogonal chemical proteomics. We demonstrated that ortho-hydroxyl substituent is required for efficient azobenzene-bond cleavage and show that these cleavable affinity tags can be used to identify newly synthesized proteins in bacteria targeted by amino acid chemical reporters as well as their sites of modification on endogenously expressed proteins. The azobenzene-based affinity tags are compatible with in-gel, in-solution, and on-bead enrichment strategies and should afford useful tools for diverse bioorthogonal proteomic applications.

Figure 1. Schematic of selective enrichment and elution of alkyne/azide-labeled proteins or peptides using diazobenzene-based cleavable affinity tags for bioorthogonal proteomics

In this two-step labeling approach, target proteins labeled with alkyne or azide-functionalized chemical reporters/probes can be selectively reacted with CuAAC reagents for detection and identification. For example, alkyne-modified proteins/peptides can be reacted with azido-diazo-biotin (1) and enriched at the protein level and eluted with Na₂S₂O₄ for gel-based proteomics or subjected to on-bead protease digestion and LC-MS/MS analysis. Alternatively, tagged peptides can be enriched and cleaved with Na₂S₂O₄ for LC-MS/MS analysis. See also Figure S1.

Figure 2. Synthesis of the second-generation diazobenzene-based cleavable affinity tags (3–6)

(a) 2-azidoethyl tosylate, K₂CO₃, DMF, 0 °C then rt, 85 %; (b) LiOH, THF/H₂O (v/v = 1/1), pH = 12.0, 90%; (c) i. oxalyl chloride, cat. DMF, CH₂Cl₂; ii. biotin-PEG-NH₂, Et₃N, CH₂Cl₂, 55 %; (d) propargyl bromide, K₂CO₃, DMF, rt, 70 %; (e) LiOH, THF/H₂O (v/v = 1/1), pH = 12, 90%; (f) i. oxalyl chloride, cat. DMF, CH₂Cl₂; ii. biotin-PEG-NH₂, Et₃N, CH₂Cl₂, 39 %; (g) resorcinol, 2-azidoethyl tosylate, EtOH, KOH_(aq), reflux, 60%; (h) i. methyl 4-aminobenzoate, 6 N HCl_(aq), NaNO₂, 0 °C, 15 min; ii. K₂CO₃, THF, pH 8.0, 0 °C then rt, 5%; (i) LiOH, THF/H₂O (v/v = 1/1), pH = 12.0, >95%; (j) N-hydroxysuccinimide, DCC, THF, 2.5 h; (k) biotin-PEG-NH₂, DMF, 4 h, 65% over 2 steps; (l) Br₂, AcOH, 84%; (m) i. 4-aminobenzoic acid, 6 N HCl_(aq), NaNO₂, 0 °C, 25 min; ii. K₂CO₃, THF, pH 8.0, 0 °C then rt, 8–15%; (n) N-hydroxysuccinimide, DCC, THF, 3 h; (o) biotin-PEG-NH₂, DMF, 4 h, 53% over 2 steps. See also Figure S4.

Figure 3. (A) Chemical structures of 2-aminooctynoic acid (AOA) and azidonorleucine (ANL). (B) In-gel fluorescent profiling of Met- AOA- and ANL-metabolically-labeled bacterial proteomes

MetRS-NLL S. typhimurium was labeled with 1 mM Met (negative control), AOA or ANL for 1 hour. The cell lysates were subjected to CuAAC with azido-rhodamine (Charron et al., 2009) and analyzed by in-gel fluorescence scanning. Coomassie blue gel shows proteins were equally loaded.

Figure 4. HPLC analysis of the diazobenzene reduction efficiencies for affinity tags 1 and 3–6

Each diazobenzene-based affinity tag (2 μL from 5 mM stock solution, final concentration = 0.1 mM) was treated with 100 μL of freshly-made Na₂S₂O₄ (in PBS, pH 7.4) of the indicated concentrations. At the described time points, the reaction solution was immediately injected into analytical reversed-phase HPLC. HPLC analysis was conducted with H₂O/CH3CN: 90%/10% to 15%/85% over 20 min. (A) Schematic of reduction of diazobenzene-based affinity tags by Na₂S₂O₄. (B) azido-diazo-biotin (1). (C) azidoethoxy-diazo-biotin (3). (D) alkynylmethoxy-diazo-biotin (4). (E) ortho-hydroxyl-azidoethoxy-diazo-biotin (5). (F) Treatment of tag 4 with 300 mM Na₂S₂O₄ solution of various urea/thiourea concentrations. The cleavage efficiencies enhanced when the amounts of thiourea increased in the cleavage solution. The cleavage efficiency difference between 6 M urea/2 M thiourea and 8M urea indicates that thiourea plays the major role in cleavage efficiency enhancement. (G) bromo-azido-diazo-biotin (6). See also Figure S3.

Figure 5. SDS-PAGE analysis of Na₂S₂O₄-protein elution efficiencies of compounds 1, 3, 5 and 6-tagged proteins

Coomassie blue gel images show the elution profiles of metabolically-labeled proteins which were reacted with either tag 1, 3, 5 or 6 via CuAAC. Biotinylated proteins were enriched with streptavidin beads from 1.5 mg total cell lysates and then eluted with 25 mM Na₂S₂O₄ (in 1% SDS, pH 7.4, for tags 1, 5 and 6) and 300 mM Na₂S₂O₄ (in 1% SDS, pH 7.4, for tag 3). E1: the first elution fraction, E2: the second elution fraction, E3: the third elution fraction, B: the eluant from boiling the streptavidin beads in 4% SDS buffer/10% β-mercaptoethanol/1×LDS for 10 min. Each elution fraction represents 30 min Na₂S₂O₄-treatment. The beads were washed twice between each elution using the cleavage buffer containing no Na₂S₂O₄. See also Figure S5.

Figure 6. Overview of the proteomic analysis of AOA/ANL-labeled S. typhimurium proteins using diazobenzene-based cleavable affinity tags

(A) Comparison of tag 1 and tag 6 in the numbers of their MS-identified AOA-labeled proteins via in-gel digestion approach and the comparison of these data to the previously published data (tag 1*) (Grammel et al., 2010). Starting with 4 mg of 1 mM Met/AOA-metabolically labeled cell lysates, the in-gel digestion approach gave total 456 protein hits for either tag 1 or tag 6. (B) Comparison of tag 1 and tag 4 in the numbers of their MS-identified AOA- and ANL-metabolically labeled proteins and unique modified peptides via in-solution digestion approach. Starting with 10 mg of 1 mM Met/AOA and Met/ANL-metabolically labeled cell lysates, the in-solution digestion approach yields total 185 unique modified peptides accounted for 73 proteins for tag 1 and total 128 unique modified peptides accounted for 65 proteins for tag 4. (C) Comparison of on-beads digestion approach and in-solution digestion approach in the numbers of their MS-identified AOA-metabolically labeled proteins and unique modified peptides using 6 as the affinity tag. Starting with 2 mg of 1 mM Met/AOA-metabolically labeled cell lysates, the in-solution digestion approach yields total 44 unique modified peptides accounted for 16 proteins whereas the on-beads digestion approach yields total 136 proteins as well as 36 unique modified peptides. See also Figure S5D and Tables S1–S5.

Figure 7. Selected MS/MS spectra of the modified universal stress protein G peptides (aa 70–83)

AOA/ANL-metabolically-modified proteins were clicked to tag 1, 4 or 6 via CuAAC and then processed via in-solution proteomic strategy. Selected MS/MS spectra were derived from the results of MS analysis of those streptavidin-enriched and Na₂S₂O₄-eluted modified peptides. (A) MS/MS spectrum of the modified peptide LQTM*VGHFSIDPSR (M* = AOA + tag 1/CuAAC/Na₂S₂O₄-cleavage adduct, the molecular weight of M* = 315). (B) MS/MS spectrum of the modified peptide LQTM*VGHFSIDPSR (M* = ANL + tag 4/CuAAC/Na₂S₂O₄-cleavage adduct, the molecular weight of M* = 301). (C) MS/MS spectrum of the modified peptide LQTM*VGHFSIDPSR (M* = AOA + tag 6/CuAAC/Na₂S₂O₄-cleavage adduct, the molecular weight of M* = 393). See also Tables S2–S4.