Synthesis and Evaluation of N-Arylsulfonylated Succinimides as Activity-Based Probes

Abstract

Activity-based protein profiling (ABPP) technology has served as a powerful platform for studying proteins for more than two decades. However, the further growth of this field depends on the development of new probe structures to expand the proportion of the proteome that can be studied using these methods. Inspired by previous reports of succinimide-containing covalent inhibitors for proteases, we synthesized a panel of potential probe structures with a succinimide reactive group and a terminal alkyne tag suitable for subsequent azide-alkyne click chemistry. Members of this panel with an N-arylsulfonyl linker produce labeling of both purified serine proteases as well as proteins in complex cellular lysates. We found that one of these probes labels the human rhomboid protease RHBDL2 at low micromolar concentrations and can be competed with active-site inhibitors. Our studies establish succinimide as a new reactive group for the development of activity-based probes and offer a new chemical tool for studying a class of enzymes with limited functional characterization.

Key words

chemical probes - succinimides - click chemistry - activity-based protein profiling

Since its development more than twenty-five years ago, activity-based protein profiling (ABPP) technology has served as a powerful strategy for studying proteins.[1] ABPP methods have facilitated the identification of aberrant protein activity in disease, the functional annotation of orphan proteins, and the development of enzyme inhibitors.[2] Central to the use of this technology is the design and synthesis of suitable activity-based probes, molecules that covalently label one or more proteins of interest for detection and/or isolation. An activity-based probe typically consists of a reactive group that engages the target protein(s), a linker, and either a tag that can be used for detecting protein engagement or a latent reactive handle for later incorporation of a tag (Figure [1]A). Though ABPP technology has been employed for many protein classes, ranging from cysteine proteases[3] to deubiquitinating enzymes,[4] the development of new probe structures can drive further advances in this methodology.

One of the original protein classes studied by ABPP technology is the superfamily of serine hydrolases.[5] Several reactive groups have furnished suitable activity-based probes for serine hydrolases. Fluorophosphonates have notably found widespread use as they have been shown to covalently react with a large subset of this superfamily.[6] These structures, however, do not provide complete coverage of the family and can be difficult to synthesize. Other reactive groups that have furnished activity-based probes for serine hydrolases include β-lactones,[7] isocoumarins,[8] and carbamates[9] among others.

Figure 1 Succinimide as a potential reactive group for an activity-based probe. (a) General structure of an activity-based probe with either a reporter tag or a latent reactive handle; (b) Proposed covalent engagement of a serine nucleophile by succinimide; (c) Structures of previously reported succinimide-containing inhibitors for human leukocyte elastase[12] and PARL,[11] respectively.

Succinimides have been proposed to serve as covalent inhibitors of serine hydrolases through addition of the nucleophilic serine into one of the imide carbonyls (Figure [1]B).[10] The succinimide ring is then proposed to open, resulting in an ester linkage with the potential for further reactivity depending on the nature of the substituent on nitrogen. We previously reported that N-arylsulfonylated succinimides could serve as inhibitors for the intramembrane serine protease PARL (Figure [1]C).[11] Related succinimide-containing structures have been developed to serve as covalent inhibitors of soluble serine proteases, including human leukocyte elastase and α-chymotrypsin.[12] Though 4-oxo-β-lactams have been described as activity-based probes for serine hydrolases,[13] the related succinimides have not yet been studied as potential activity-based probes to the best of our knowledge. For this study, we synthesized a panel of five alkyne-functionalized succinimide-containing probe structures and investigated their ability to engage both purified soluble serine proteases and proteins in complex mammalian proteomes.

Synthesis of probes : In designing our panel of probes, we sought to investigate a variety of linkers between the reactive group of the probe (the succinimide) and the terminal alkyne tag, a biorthogonal reactive handle for later incorporation of a fluorophore for visualization.[14] We generated a panel of five initial probes (Scheme [1]): one with a simple methylene linker, one with a longer polyethylene glycol (PEG) connector, and three structures with N-arylsulfonyl units inspired by our work developing PARL inhibitors.[15]

Scheme 1 Synthesis of a panel of succinimide-containing probes 1–5. Probes 1 and 2 were synthesized through a one-step process from the corresponding amines. Probes 3, 4, and 5 were synthesized through 2–4 step routes. Reagents and conditions: (a) aq. NH₄OH, THF, 23 °C (82% yield); (b) Pd(PPh₃)₂Cl₂, CuI, (triethylsilyl)acetylene, NEt₃, 40 °C (97% yield); (c) TBAF, THF, 23 °C (65% yield); (d) propargyl bromide, K₂CO₃, DMF, 23 °C (48% yield); (e) hept-6-yn-1-ol, PPh₃, DIAD, THF, 23 °C (28% yield); (f) 1. THF, NEt₃, 50 °C, 2. Ac₂O, 23 °C (29–42% yield).

Figure 2 Reactivity of the succinimide probes with soluble serine proteases. (a) Workflow for the gel-based ABPP experiments; (b) ABPP gels for trypsin and α-chymotrypsin treated with either the indicated probe at 100 μM or the DMSO vehicle for 30 min, followed by reaction with Rh-N₃ for 1 h; (c) ABPP gels for trypsin and α-chymotrypsin treated with the indicated concentration of compound 4 or the DMSO vehicle for 30 min, followed by reaction with Rh-N₃ for 1 h. All gels were performed in duplicate with a representative image shown.

Probes 1 and 2 were accessed in a single step from the reaction of commercially available amines with succinyl chloride using previously reported conditions.[16] Probe 3, which features an ethynyl-substituted N-arylsulfonyl substituent, was accessed from 4-bromobenzenesulfonyl chloride in four steps. After formation of the primary sulfonamide by addition of ammonia to the sulfonyl chloride, the ethynyl substituent was installed using a Sonogashira coupling with (triethylsilyl)acetylene followed by removal of the silyl protecting group with tetra-n-butylammonium fluoride (TBAF). A one-pot procedure with sequential treatment of the primary sulfonamide with succinic anhydride followed by acetic anhydride furnished the final probe structure.[17] The second N-arylsulfonylated compound, probe 4, contains a short ethereal linkage between the aromatic ring and the terminal alkyne. This group was installed by alkylation of 4-hydroxybenzenesulfonamide with propargyl bromide in the presence of potassium carbonate. The final compound, probe 5, contains a longer linker that was generated through a Mitsunobu reaction between 4-hydroxybenzenesulfonamide and hept-6-yn-1-ol. Though relatively low yielding, this method allowed us to access our desired compound to obtain structures with a variety of linker lengths. An identical set of conditions to those employed for probe 3 were used to install the succinimides of probes 4 and 5. While compound 1 has been previously described in the literature,[18] the other four structures 2–5 have not been previously reported to the best of our knowledge.

Probe reactivity with purified serine proteases : After accessing each of the probe structures, we examined their ability to engage two purified serine proteases, trypsin and α-chymotrypsin. For our initial experiments, a solution of each protein (25 μg mL^–1) was treated with 100 μM of the desired probe for 30 min, followed by reaction with an azide-functionalized fluorophore (Rh-N₃) in the presence of CuSO₄, sodium ascorbate, and the ligand THPTA for 1 h. The samples were then quenched and separated by SDS-PAGE prior to an in-gel fluorescence measurement (Figure [2]A).

Under these conditions, we observed labeling of both trypsin and α-chymotrypsin with probes 3, 4, and 5, which feature an arylsulfonyl linker between the succinimide and the terminal alkyne (Figure [2]B). In contrast, no labeling was observed with probes 1 and 2, which contain propargyl and polyether linkers, respectively. To further probe the observed labeling events, we examined a range of concentrations for probe 4 against both trypsin and α-chymotrypsin (Figure [2]C). For both proteins, we observed that the strength of the fluorescent band decreased with treatment with progressively lower concentrations of 4, consistent with a concentration-dependent labeling event.

Figure 3 A pyrrolidine-containing analogue does not engage trypsin or α-chymotrypsin. Structure of the control compound 10 and ABPP gels for trypsin and α-chymotrypsin treated with the indicated concentration of 3, 10, or the DMSO vehicle for 30 min, followed by reaction with Rh-N₃ for 1 h. All gels were performed in duplicate with a representative image shown.

To confirm that the succinimide was essential for the observed labeling events, we synthesized an analogue of 3 with a pyrrolidine ring replacing the succinimide (compound 10) in three steps (Figure [3]).[19] This control compound lacks the proposed site of nucleophilic attack but is otherwise identical to the probe. When examined in parallel with probe 3, we failed to observe labeling of either trypsin or α-chymotrypsin by compound 10 at the same concentrations that produced robust labeling of these enzymes by 3. These results offer compelling evidence that the carbonyls of the succinimide are critical for the observed labeling events.

Probe reactivity in complex proteomes : Having investigated the reactivity of the probes with purified proteins, we then shifted to studying their behavior in complex proteomes. We prepared the membrane fraction of HEK293T cells by ultracentrifugation of the cell lysates and resuspension in phosphate-buffered saline (PBS). We first examined the global reactivity of probe 5 in this complex mixture by testing a range of concentrations from 250 to 0.1 μM. We observed broad reactivity of 5 across the proteome at the highest concentrations, but a limited number of visible bands at low micromolar concentrations and below (Figure [4]A). This result indicated that the probe could potentially achieve more selective binding if employed at low micromolar concentrations.

Having previously observed inhibition of a rhomboid protease by an N-arylsulfonylated succinimide,[11] we investigated whether we could observe labeling of a rhomboid protease with probe 5. We overexpressed human RHBDL2 (hRHBDL2), an enzyme implicated in cell migration and proliferation,[20] in HEK293T cells and generated the membrane fraction of the transfected cells. The membrane fraction was treated with probe 5 for 1 h at 1 μM, followed by reaction with Rh-N₃ for 1 h. After SDS-PAGE separation of the labeled proteome, we observed a new band at the expected molecular weight of hRHBDL2 (~34 kDa) that was not present in a mock-transfected control (Figure [4]B). To confirm that the active site serine was involved in the labeling event, we generated a mutant of hRHBDL2 with the nucleophilic serine (S187) replaced by alanine. In the presence of 1 μM 5, we observed that, despite similar expression levels as confirmed by Western blot, labeling was dramatically reduced for the mutant compared to the wild-type enzyme. This result suggests that the nucleophilic serine is primarily responsible for the probe labeling observed at this concentration. Interestingly, when we tested the probe at higher concentrations, we observed more pronounced labeling of the mutant, indicating that 5 may have additional sites of reactivity outside of the active site when applied at increased concentrations (Supporting Figure [1]).

Figure 4 Reactivity of probe 5 in a complex proteome. (a) ABPP gel of the membrane fraction of HEK293T cells treated with the indicated concentrations of 5 for 1 h, followed by reaction with Rh-N₃ for 1 h; (b) ABPP gel and Western blot for the membrane fractions of HEK293T cells transfected with empty vector (‘Mock’), wild-type human RHBDL2 (hRHBDL2), or the S187A mutant of hRHBDL2. The fractions were treated with 1 μM 5 for 1 h, followed by reaction with Rh-N₃ for 1 h; (c) Competitive ABPP gel and Western blot for the membrane fractions of hRHBDL2-transfected HEK239T cells treated with the indicated concentration of MAFP for 30 min, followed by 1 μM 5 for 30 min, then reacted with Rh-N₃ for 1 h. All gels were performed in duplicate with a representative image shown.

By using probe 5 at a lower concentration, we were then able to perform a competitive ABPP experiment. The hRHBDL2-transfected membrane fraction was first treated with a panel of serine protease inhibitors for 30 min, followed by treatment with 1 μM probe 5 for 30 min prior to reaction with Rh-N₃ (Supporting Figure [2]). We found that the fluorophosphonate MAFP and 3,4-dichloroisocoumarin were both able to compete probe labeling, the latter of which has previously been described as an hRHBDL2 inhibitor.[21] In further experiments with MAFP, we were able to successfully compete labeling by the probe in a concentration-dependent manner (Figure [4]C). Successful competition of hRHBDL2 labeling indicates that the active site-directed inhibitor successfully prevents probe engagement. These findings suggest that a competitive ABPP platform with probe 5 could be used to screen for new inhibitors of this enzyme.

Conclusions and future directions : Inspired by the use of succinimides in the structures of covalent serine protease inhibitors, we synthesized a panel of succinimide probes with a variety of linkers. Intriguingly, among this initial panel of structures, only the probes with N-arylsulfonyl linkers produce labeling of the proteins that we investigated. Future studies could investigate a broader range of linkers to determine what aspects of the linker are essential to produce a sufficiently reactive probe. We also established that the succinimide carbonyls, previously proposed to be the site of covalent addition by nucleophiles, are critical for the observed probe labeling events.

N-Arylsulfonylated succinimides appear to be broadly reactive when applied at higher concentrations (≥100 μM) but label more selectively at low micromolar concentrations. At these lower concentrations, we were able to observe labeling of human RHBDL2 with one of our probes in a manner that is largely dependent on the presence of the active site serine nucleophile. Engagement of the active site of hRHBDL2 was further supported by successful competition of probe labeling by the active site inhibitor MAFP.

Notably, increased labeling of a mutant lacking this serine nucleophile at higher probe concentrations suggests that the probe may react with additional sites outside of the active site when present in larger amounts. While beyond the scope of this study, profiling of the reactivity preferences for the succinimide probes among biological nucleophiles using mass spectrometry-based proteomics could provide insight into their selectivity for serine versus other amino acids.[22]

Collectively, our studies offer a new reactive group to consider in the design of activity-based probes. As chemical tools to study the rhomboid proteases remain limited, we have described a first-generation activity-based probe for RHBDL2 that is suitable for gel-based competitive ABPP experiments to enable inhibitor discovery. Further optimization of this probe could provide a tool for selectively monitoring the activity of RHBDL2 in various cell types to provide deeper insight into the physiological functions of this enzyme.

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

The authors wish to thank Jay Aghanya and Charles Wilber for their preliminary work on this project as well as members of the Oberlin Chemistry and Biochemistry department for helpful discussions.

• References and Notes

• 1a Cravatt BF, Wright AT, Kozarich JW. Annu. Rev. Biochem. 2008; 77: 383
• 1b Activity-Based Protein Profiling . Cravatt BF, Hsu K.-L, Weerapana E. Springer; Cham: 2019
• 2a Nomura DK, Dix MM, Cravatt BF. Nat. Rev. Cancer 2010; 10: 630
• 2b Barglow KT, Cravatt BF. Nat. Methods 2007; 4: 822
• 2c Niphakis MJ, Cravatt BF. Annu. Rev. Biochem. 2014; 83: 341
• 3 Kato D, Boatright KM, Berger AB, Nazif T, Blum G, Ryan C, Chehade KA. H, Salvesen GS, Bogyo M. Nat. Chem. Biol. 2005; 1: 33
• 4 Love KR, Pandya RK, Spooner E, Ploegh HL. ACS Chem. Biol. 2009; 4: 275
• 5a Simon G, Cravatt BF. J. Biol. Chem. 2010; 285: 11051
• 5b Faucher F, Bennett JM, Bogyo M, Lovell S. Cell Chem. Biol. 2020; 27: 937
• 6 Bachovchin DA, Ji T, Li W, Simon GM, Blankman JL, Adibekian A, Hoover H, Niessen S, Cravatt BF. Proc. Natl. Acad. Sci. U. S. A. 2010; 107: 20941
• 7a Böttcher T, Sieber S. Angew. Chem. Int. Ed. 2008; 47: 4600
• 7b Lehmann J, Cheng T.-Y, Aggarwal A, Park AS, Zeler E, Raju RM, Akopian T, Kandror O, Sacchettini JC, Moody DB, Rubin EJ, Sieber SA. Angew. Chem. Int. Ed. 2018; 57: 348
• 8a Haedke U, Götz M, Baer P, Verhelst SH. L. Bioorg. Med. Chem. 2012; 20: 633
• 8b Voskya O, Vinothkumar KR, Wolf EV, Brouwer AJ, Liskamp RM. J, Verhelst SH. L. Proc. Natl. Acad. Sci. U. S. A. 2013; 110: 2472
• 9a Chang JW, Cognetta AB, Niphakis MJ, Cravatt BF. ACS Chem. Biol. 2013; 8: 1590
• 9b Cognetta AB, Niphakis MJ, Lee H.-C, Martini ML, Hulce JJ, Cravatt BF. Chem. Biol. 2015; 22: 928
• 10 Martyn D, Moore N, Abella A. Curr. Pharm. Des. 1999; 5: 405
• 11 Parsons WH, Rutland NT, Crainic JA, Cardozo JM, Chow AS, Andrews CL, Sheehan BK. Bioorg. Med. Chem. Lett. 2021; 49: 128290
• 12a Groutas WC, Brubaker MJ, Stanga MA, Castrisos JC, Crowley JP, Schats EJ. J. Med. Chem. 1989; 32: 1607
• 12b Groutas WC, Venkataraman R, Brubaker MJ, Stanga MA. Biochemistry 1991; 30: 4132
• 13a Ruivo EF. P, Gonçalves LM, Carvalho LA. R, Guedes RC, Hofbauer S, Brito JA, Archer M, Moreira R, Lucas SD. ChemMedChem 2016; 11: 2037
• 13b Carvahlo LA. R, Ross B, Fehr L, Bolgi O, Wöhrle S, Lum KM, Podlesainski D, Vieira AC, Kiefersauer R, Félix R, Rodrigues T, Lucas SD, Groß O, Geiss-Friedlander R, Cravatt BF, Huber R, Kaiser M, Moreira R. Angew. Chem. Int. Ed. 2022; 61: e202210498
• 14 Speers AE, Cravatt BF. Chem. Biol. 2004; 11: 535
• 15 Full details regarding the experimental procedures and characterization data for the synthesized compounds are available in the Supporting Information.
• 16 Zhang Y.-J, Shen L.-L, Cheon H.-G, Xu Y.-N, Jeong J.-H. Arch. Pharm. Res. 2014; 37: 588
• 17 Nocentini A, Ferraroni M, Carta F, Ceruso M, Gratteri P, Lanzi C, Masini E, Supuran CT. J. Med. Chem. 2016; 59: 10692
• 18 Marson CM, Khan A, Porter RA. J. Org. Chem. 2001; 66: 4771
• 19 Full details regarding the synthesis of compound 10 are available in the Supporting Information.
• 20a Düsterhöft S, Künzel U, Freeman M. Biochim. Biophys. Acta, Mol. Cell Res. 2017; 1864: 2200
• 20b Chen S, Cai K, Zheng D, Liu Y, Li L, He Z, Sun C, Yu C. Cell Death Dis. 2022; 13: 945
• 21 Cheng T.-L, Wu Y.-T, Lin H.-Y, Hsu F.-C, Liu S.-K, Chang B.-I, Chen W.-S, Lai C.-H, Shi G.-Y, Wu H.-L. J. Invest. Dermatol. 2011; 131: 2486
• 22a Adam GC, Cravatt BF, Sorensen EJ. Chem. Biol. 2001; 8: 81
• 22b Weerapana E, Simon GM, Cravatt BF. Nat. Chem. Biol. 2008; 4: 405

Publication History

Received: 10 March 2025

Accepted after revision: 10 April 2025

Article published online:
26 May 2025