Direct intracellular delivery of benzene diazonium ions as observed by increased tyrosine phosphorylation

Abstract

A challenge within the field of bioconjugation is developing probes to uncover novel information on proteins and other biomolecules. Intracellular delivery of these probes offers the promise of giving relevant context to this information, and can serve as hypothesis generating tools within complex systems. Leveraging the utility of triazabutadiene chemistry, herein we discuss the development of a probe that undergoes reduction-mediated deprotection to rapidly deliver a benzene diazonium ion (BDI) into cells. The intracellular BDI resulted in an increase in global tyrosine phosphorylation levels. Seeing phosphatase dysregulation as a potential source of this increase, a tyrosine phosphatase (PTP1B) was tested and shown to be both inhibited and covalently modified by the BDI. In addition to the expected azobenzene formation at tyrosine side-chains, key reactive histidine residues were also modified.

Graphical Abstract

INTRODUCTION

Selectivity in biological systems comes from a complex interplay of location, interactions and reactivity. Covalent small molecule probes offer great potential in the development of chemical tools to study intracellular proteins.^1,2 A challenge within this area is in garnering selectivity associated with location. Oftentimes, in order to gain accessibility to intracellular proteins, small molecule labeling strategies rely on working with cell-lysate. While powerful, key contextual interactions associated with localization within the cell are inevitably lost during lysis (Figure 1). Tyrosine plays a pivotal role at the interface of these interactions, and as such probes that can modify tyrosine are of value.^3–6 One such class, aryl diazonium ions, are best known for their selective reactivity with the electron-rich aromatic tyrosine side chain.^7,8 While these species are primarily known for their reactivity with tyrosine, histidine was historically the first to be observed although its reactivity is not as ubiquitous across the range of electronically diverse aryl diazonium ions.^9,10 Like tyrosine, histidine is often found at interfaces,^11,12 but it also plays a critical role in enzymatic active sites often as a base or metal binder.¹³ As with other small molecule probes who are “always on,” aryl diazonium ions suffer from a lack of deliverability. To address this limitation, herein we report a novel chemical probe that selectively releases benzene diazonium ions (BDI) intracellularly, opening the door for a better understanding of the state of a cell pre-lysis.

Figure 1.

Previous bioconjugation work with benzene diazonium ions but has been limited primarily to in vitro use. This work focuses on developing a delivery method for benzene diazonium ions so that they may be used to study proteins in their native environment and the biochemical processes of live cells.

Aryl diazonium ions have undergone a renaissance in the bioconjugation literature,^4,14–19 but their chromophoric properties enabled them to be one of the first reliable methods to modify proteins dating back to the early 20^th century.^10,20,21 While useful, many aryl diazonium ions suffer from the requirement of generation immediately prior to the experiment due to their short half-lives in solution and general instability. Recognizing the promise of aryl diazonium ions as robust small molecule covalent probes, the triazabutadiene was developed as a new biochemical tool for their generation.²² At physiological pH upon protonation at the N3 position (Figure 2a), the triazabutadiene is protonated to release a BDI and a guanidinium side product. This circumvents the challenges of BDI instability. Further work focused on controlling the reactivity of the triazabutadiene by varying the groups at the R1 and R2 positions to control the rate of BDI release.²³ Simply governing the pH-release of the BDI does not afford the level of selectivity that we desired. In order to better dictate the location in which a BDI is released we sought a second level of control.

Figure 2.

a. The triazabutadiene releases a BDI under mild conditions at physiological pH. The N1 position of the triazabutadiene can be protected via a covalent modification to prevent BDI release. The protected triazabutadiene can then be deprotected to rescue triazabutadiene reactivity and BDI release. b. Azo-tyrosine adducts are unable to be phosphorylated. Conversely, phosphorylated tyrosine residues cannot react with BDIs. BDI conjugation to tyrosine is inversely related to tyrosine phosphorylation.

We found that formylating the N1 position of the triazabutadiene protected the release of the BDI physiological pH and acidic conditions.²⁴ Upon exposure to a high pH environment, the triazabutadiene is deprotected and once again able to release a BDI (Figure 2a). While the base-labile protection approach has utility in select settings, the general concept of protecting the N1 position provides a strategy by which we can obtain desired selectivity and deliverability in a range of settings (Figure 2a). The study herein reports on utilizing the triazabutadiene protection strategy to deliver a BDI to determine its in cellulo effects. Taking advantage of the high intracellular reducing potential,²⁵ a reduction-sensitive group was envisioned to offer spatial control of the probe with uncaging of the triazabutadiene initiated only in a reducing intracellular environment. To the best of our knowledge, this is the first report of a triazabutadiene being utilized in cellular experiments, and the first report of a BDI being selectively released intracellularly.

MATERIALS AND METHODS

Materials.

4G10 Platinum and Anti-Phosphotyrosine Antibody (05-1050) were obtained from EDM Millipore. Rabbit-β-Tubulin antibody (H-235), Phosphatase Inhibitor Cocktail A, and Phosphatase Inhibitor Cocktail B were obtained from Santa Cruz Biotechnologies. 6, 8-Difluoro-4-Methylumbelliferyl Phosphate (DiFMUP), Alexa Fluor 555 goat anti-rabbit IgG (A-21428), and Alexa Fluor 647 goat anti-mouse IgG (A-21236) were obtained from Thermo Fisher Scientific (Invitrogen). GAPDH (14C10) Rabbit mAb was obtained from Cell Signaling Technology. Mammalian Protein Extraction Reagent (M-PER) and Micro BCA Protein Assay Kit were obtained from Thermo Fisher Scientific. Recombinant human PTP1B and PP1 were obtained from Novus Biologicals.

Synthesis of (2-((chlorocarbonyl)oxy)ethyl) methanesulfonothioate (5).

To a flame dried flask was added 100 mg of crushed 4 Å molecular sieves. Then was added a solution K₂CO₃ (1.3 mmol) in toluene (4 mL) under argon. The solution was cooled to −10 °C and (2-hydroxyethyl) methanesulfonothioate (1.1 mmol) was added to the reaction vessel slowly. Then a 20% solution of phosgene in toluene (1 mmol) was added dropwise to the solution over 10 minutes. The reaction was allowed to stir for 30 minutes at −10 °C, and then at room temperature for 4 h under argon. After 4 h, argon was bubbled through the reaction mixture for 5 minutes in a closed hood with an outlet in the septa to remove excess phosgene. The reaction was filtered over a pad of MgSO₄, and washed with ether. The resulting filtrate was evaporated to dryness to yield 5, which was taken forward and used without further purification.

Synthesis of (E)-3-(tert-butyl)-1-methyl-2-(3-((2-((methylsulfonyl)thio)ethoxy)carbonyl)-3-phenyltriaz-1-en-1-yl)-1H-imidazol-3-ium chloride (3). A flame dried and vacuum evacuated flask under argon was charged with 4Å molecular sieves and dichloromethane. To this solution was added 5 (0.8 mmol). This solution was allowed to stir under argon for 5 minutes. To this was added 2 (0.07 mmol) in one portion at room temperature. The reaction was left to stir under argon for 12 h, after which time it was filtered and the filtrate was concentrated down to yield a yellow solid. Purification of 3 involved a silica column with 10% MeOH/CH₂Cl₂. After which the resulting product was dissolved in CH₂Cl₂ and washed 3x with an aqueous 0.1% TFA solution. The CH₂Cl₂ layer was evaporated to dryness to provide 3 as a yellow solid (0.021 g, 67% yield).

In cellulo global tyrosine phosphorylation assays.

HEK 293T cells were maintained in complete media (90% DMEM, 10% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin and 2.5 μg/mL amphotericin B). Cells were plated 24 h prior to treatment. 24 h post cell plating, cells were treated with compounds or DMSO for a specified amount of time. Cells were incubated for the indicated times with appropriate compounds and then lysed with MPER supplemented with proteasome and phosphatase inhibitors for 20 mins at 4 °C while gently agitating. Cell lysates were then centrifuged at 14,000 rcf for 10 mins at 4 °C. Protein concentration of the supernatant was quantified using the BCA reagent, and 30 μg of total protein per well was loaded on SDS-PAGE gels. Proteins were transferred onto PVDF membranes and blocked using 5% BSA in TBST for 1 h at room temperature. Membranes were probed with primary antibodies of interest overnight at 4 °C while gently agitating. Membranes were then washed and probed with secondary antibody for 1 h at room temperature. Membranes were then imaged using a BioRad ChemiDoc MP Imaging System.

In vitro Phosphatase Assays.

All DiFMUP fluorogenic-based phosphatase assays were carried out at least in duplicate. All assays were carried out in a final volume of 100 μL. Enzymes (PP1 or PTP1B) were incubated in reaction buffer (50 mM Tris solution pH 7, 200 μM MnCl2, 2 mM DTT, 0.05% (v/v) Tween-20, and 125 μg/mL BSA) with DMSO only or with compounds at specific times before the addition of 100 μM DiFMUP. DiFMUP fluorescence over time was measured using a BMG LABTECH CLARIOstar Plus microplate reader. Samples were excited at 358 nm and emission scans recorded from 420–470 nm with a maximum emission at 448 nm.

Mass Spectrometry Analysis of BDI Conjugation.

Experiments were carried out in duplicate in a final volume of 100 μL. 1 μg of PTP1B was incubated in reaction buffer (50 mM Tris solution pH 7, 200 μM MnCl₂, 2 mM DTT, 0.05% (v/v) Tween-20, and 125 μg/mL BSA) with controls of DMSO only or replicates of 2. Samples were trypsin-digested, purified as previously described and subsequently analyzed by LC-MS/MS as previously described.

RESULTS AND DISCUSSION

Given its history in bioconjugation we recognized the BDI was promising to use as a covalent probe, but were presented with the following challenges: 1) delivering the BDI intracellularly and 2) determining the cellular effects of BDI treatment. First, we sought a metric to determine if our probe was able to enter the cytosol and release its reactive payload. It is known that aryl diazonium ions can be used to probe dephosphorylation of substrates, with the premise being that phosphorylated phenols are unreactive, but become reactive upon dephosphorylation.²⁶ Additionally, previous work has shown derivatized tyrosine residues have diminished-to-no reactivity, as such, a phenol that reacts with a BDI to form an azobenzene adduct would be potentially resistant to phosphorylation (Figure 2b).²⁶ Given this varying reactivity as a function of tyrosine phosphorylation, we hypothesized that global tyrosine phosphorylation was a viable output for measuring BDI release in the cell. The expectation at the onset was that global tyrosine phosphorylation, as measured by a phosphotyrosine-specific antibody, should decrease as more BDI forms azo-adducts with proteins.

Delivery of the benzene diazonium ion (BDI)

As discussed above, the triazabutadiene scaffold (Figure 2a) was an attractive delivery system because the BDI would be generated under mild conditions compatible with cells.²² For our study we chose tert-butyl/methyl triazabutadiene 2 as the half-life of the molecule was shown to be < 2 minutes at pH 8.²³ This was ideal because the BDI release is much faster than our phosphorylation assay and the timing of the release of the BDI was dependent upon the deprotection event that would liberate 2 in solution. Indeed, when working with variants of 2 the addition of the triazabutadiene into neutral water can be likened to adding freshly made aryl diazonium salts.²⁷ The reactivity of BDI suggests that it is likely able to react with membrane proteins and its small, positively charged nature suggests that it should also be able to pass through the cell’s membrane. To understand the background reactivity of a non-guided probe, we sought to investigate the effect of simple addition of BDI to cells as a control (Figure 3a). We incubated various concentrations of 2 with HEK293T cells for 3 h, lysed the cells, and measured global tyrosine phosphorylation levels using a fluorescent 2° p-Tyr antibody. Expecting to see a decrease in global tyrosine phosphorylation levels we were surprised to observe that global tyrosine phosphorylation level had increased in a 2-dependent manner (Figure 3b). We repeatedly verified these experiments and observed a marked increase in global tyrosine phosphorylation at 500 μM of 2. We conducted a time-course experiment incubating HEK293T cells with 500 μM of 2 and observed the increase in global tyrosine phosphorylation signal as early as 2 h (Figure 3c).

Figure 3.

a. Hypothesized mechanism for 2 with cells. Compound 2 will release the BDI in < 1 min, which can then either react with proteins extracellularly, diffuse through the cell membrane and react with intracellular proteins, or undergo hydrolysis. b. Concentration dependence of 2 with HEK293T cells, an increase in global p-Tyr is seen at concentrations of 500 μM and above (Figure S1). c. Time dependent experiment with 500 μM 2, an increase in global p-Tyr is observed at 120 min. While BDI conjugation to histidine is known, the figure only shows tyrosine conjugation for clarity.

We hypothesized the mechanism leading to these unanticipated results was that (a) 2 was releasing the BDI, and the BDI was diffusing through the membrane and reacting intracellularly with enzymes that dephosphorylate tyrosine residues and/or (b) the BDI was reacting with extracellular proteins that initiated a signaling cascade and caused an increase in global tyrosine phosphorylation. We also speculated that because the BDI is a highly reactive species with a half-life within the time-frame of the experiments²⁸ it was possible that a considerable amount of the BDI was lost due to hydrolysis to its unreactive phenolic form (Figure 3a). While these results were promising in that the new BDI had a potential readout, we recognized that these data alone did not support that the BDI was intracellular.

We sought to design a molecule that would deliver the BDI intracellularly. We reasoned that this molecule would not only confirm if the BDI was getting past the cell membrane, but also allow for more productive intracellular BDI delivery by increasing intracellular BDI concentrations. As discussed above, we knew that formylations of the triazabutadiene scaffold could protect against BDI release (Figure 2a). However, in our previous work we had only removed these protecting groups upon exposure to high pH environment.²⁴ We adapted this strategy to our system, and designed a protecting group that would be cleaved upon exposure to an intracellular environment. We hypothesized that this reactivity would be found in a sulfonyl thiolate (analogous to a disulfide) containing protected triazabutadiene, like 3 shown in Scheme 1. The positively charged small molecule should readily pass through the cell’s membrane. Upon exposure to the high intracellular concentrations of reducing agents,²⁵ a direct reduction of the sulfur-sulfur bond is feasible, however a multistep mechanism is also possible.²⁹ The resulting thiol is then expected to undergo rapid self-immolative deprotection, as seen in similar compounds,³⁰ and unmask 2 within the cell. As with the earlier experiments, triazabutadiene 2 is expected to readily release BDI intracellularly (Figure 4a).

Scheme 1.

Synthesis of Protected Triazabutadiene, 3

Figure 4.

a. Hypothesized mechanism for 3 with cells. Intracellular delivery of 3 triggers a reduction of the disulfide mimic and subsequent deprotection, delivering 2 intracellularly, which will then release a BDI. b. Hypothesized mechanism for control compounds 6 and 7, which should both get transferred intracellularly, but should remain in their protected form, never releasing a BDI. c. Concentration dependence of 3 with HEK293T cells, increase in global p-Tyr is observed at 250 μM and 500 μM. d. Time-dependent experiment for 3 at 500 μM with HEK293 cells, increase in global p-Tyr is observed at 30 min. e. Control compounds 6 and 7 tested alongside 2 and 3, control compounds have no effect on global p-Tyr, and 3 increases global p-Tyr at 30 min, consistent with previous results.

Protected triazabutadienes are synthesized with a nucleophilic triazabutadiene and a corresponding chloroformate electrophile. The unprotected tert-butyl/methyl triazabutadiene, 2, was synthesized in two steps from tert-butyl imidazole and phenyl azide (Scheme 1a).^31–33 For the chloroformate electrophile, we started by synthesizing alcohol 4 in two steps from previously established methods,³⁴ and treated 4 with phosgene to yield chloroformate 5. Gratifyingly, the reaction between triazabutadiene 2 chloroformate 5 provided the protected triazabutadiene 3 in reasonable yield (Scheme 1b). Unlike 2, protected 3 is stable for days in neutral water. In growth media containing resorcinol as a BDI trap, 2 readily provided the corresponding azobenzene product, and 3 remained intact until dithiothreitol was added to trigger deprotection and BDI release (Figure S2).

In addition to 3, we also synthesized two control compounds (Figure 4b). Compound 6 is an ethyl carbamate protected triazabutadiene similar to that synthesized in previous work,²⁴ which can deprotect and release a BDI only after exposure to a high pH environment. This compound helps to rule out an extracellular release mechanism of 3. The second control compound, 7, is a triazabutadiene that was alkylated at the N1 position. Alkylated triazabutadiene 7 will not release a BDI, as previous work has shown that an alkylation at this position is irreversible (Figure 4b).^35,36 If the BDI, 1, is the active biochemical agent, then no effect observed on global tyrosine phosphorylation should be seen when using either of these compounds in cellular assays.

With 3 in hand, we repeated the same concentration and time-dependent experiments as with 2. To our delight, we observed 3 considerably increased global tyrosine phosphorylation at 500 μM, however we observed this increase to also be accompanied by a loss of the β-tubulin and GAPDH housekeeping proteins (Figure 4c). Additionally, upon using the 3 in time-course experiments we observed the increase in global tyrosine phosphorylation as early as 30 minutes, consistent with previously observed rates of similar self immolative compounds (Figure 4c,d).³⁰ Upon directly comparing 2,3,6 and 7, there was no effect on global phosphorylation observed for either 6 or 7 at 30 min or up to 180 mins (Figure 4e & S3) further bolstering the model of an intracellular BDI being potentially responsible for the observed increase in global tyrosine phosphorylation.

These data are consistent with the model that 3 acts as a tool for intracellular delivery of 1. Regarding the previous experiments with 2, we reasoned that 1 is also successful in reaching the intracellular space, as the same trend in increasing global tyrosine phosphorylation was seen with 2 and 3. Based on the times at which the change in phosphorylation was noted (30 min for 3 versus 2 h for 2), it appears that 3 is more efficient at intracellular delivery of 1 than simply introducing 1 to the environment around the cells. We note that at higher concentrations and prolonged treatment times, that the housekeeping proteins β-tubulin and GAPDH are greatly reduced in intensity (Figure 4c, d). The catastrophic effect on housekeeping proteins indicates some large change in cellular homeostasis, which may or may not be a desirable outcome, and will be investigated in future experiments.

Cellular effects of BDI treatment

Having established that 1 is likely being released intracellularly, we sought to better understand why we observed the effects that we did. There have been limited attempts of biological application with aryl diazonium ions, and to the best of our knowledge, this is the first report of an aryl diazonium ion being delivered directly into cells, and of the effect on global tyrosine phosphorylation. During the course of our experiments, we observed that the cell morphology had changed during the 3 h experiment and that cells appeared to be dying throughout the experiment (Figure S4). As mentioned earlier, at longer time points (>2 h), concomitant with an increase in global tyrosine phosphorylation, there was also an observed disappearance of the housekeeping proteins, β-tubulin and GAPDH (Figure 4c&d). These data suggested to us that the BDI is toxic and causes significant cellular stress. It is noteworthy that carcinogenic effects of certain aryl diazonium ions has been previously noted.^37–39

One mechanism to explain the carcinogenic effects for para-substituted aryl diazonium ions, like those found in common white mushrooms,⁴⁰ is that they can form aryl radicals which go on to arylate nucleotides and cause DNA strand damage.^41–43 Previous reports reason that this damage could be the cause of activation of AP-1, through the MAP kinase pathway as observed by antibodies targeting phosphorylated members of the pathway.⁴⁴ While a mechanism of DNA damage leading to kinase activation cannot be wholly ruled out for 1, the unsubstituted BDI has not been shown to do this kind of chemistry. Indeed, in the absence of a reductant, like copper, aryl diazonium ions lacking radical-stabilizing substituents do not undergo DNA damaging chemistry.⁴⁵ Given the large body of literature that shows aryl diazonium ions undergo protein-based chemistry, and the rapid onset of global phosphorylation, we sought to explore a protein-based rationale for why such a high level of phosphorylation was observed. We turned our attention to factors that control cellular levels of phosphorylation and hypothesized that the BDI may be directly inhibiting tyrosine phosphatases, thereby disrupting the dephosphorylation process leading to an accumulation of phosphorylated proteins.

We first turned our attention to tyrosine phosphatase PTP1B. An in vitro DiFMUP phosphatase inhibition assay was conducted measuring phosphatase activity as a function of fluorescence. As mentioned previously, after adding 2 to a solution it will generate 1 rapidly. As such, for these assays 2 was used to determine the effects the BDI had on phosphatases. We observed that treatment of PTP1B with 2 inhibits PTP1B with an IC₅₀ of 2.5 μM (Figure S5a). We next were interested to see if this inhibition activity was specific to tyrosine phosphatases or could be more general. We repeated the inhibition assay with PP1, an unrelated serine/threonine phosphatase, and observed that PP1 was also inhibited with an IC₅₀ of 3.8 μM (Figure S5a). As a control, the assay was also conducted with 3 and PP1. As expected because of the absence of the intracellular reduction trigger, there was virtually no inhibition observed at all concentrations of 3, further supporting the BDI is responsible for the observed inhibition (Figure S6a). The inhibition of phosphatases explains the observed changes in phosphorylation. It is likely that 1 is promiscuous and targets a variety of intracellular proteins. Our assay was focused on tyrosine phosphorylation, but that does not preclude other covalent modifications or effects that these reactive species could have on the cell. While phosphatase inhibition is likely a culprit, the interaction with DNA cannot be ruled out and the two pathways could act synergistically to increase phosphorylation levels.

BDI side chain interactions

Intrigued by the potential covalent inhibition of PTP1B, we subjected samples to proteomics analysis to identify the specific residues that had been modified by 1. Knowing that aryl diazonium ions react with surface tyrosine and histidine residues, and given the reactivity and large stoichiometric excess of the BDI, we expected that most, if not all, solvent exposed tyrosine and histidine residues would be covalently modified. Experiments were run with 10 μM 2 and 20 nM PTP1B, following the same conditions as the in vitro inhibition assays with PTP1B. The samples were subjected to trypsin digestion and subsequent proteomic analysis with mass spectrometry.

We first conducted a spectrum counting analysis with the Scaffold program to evaluate for the presence of modifications. Contrary to our hypothesis, we observed that many surface exposed, or ‘easily accessible’, tyrosine and histidine residues on PTP1B were not modified by the BDI. Indeed, we observed only six tyrosine and histidine residues that underwent modification: H54, H60, Y66, H94, Y176, and H214 (Figure 5). While the azobenzene was seen to be intact on some residues, others presented as a cleaved modification.^46,47 We then quantified the extracted ion abundance of these modifications using the program Progenesis QI for Proteomics.⁴⁸ We found that compared to a control sample H54, H60, and Y66 all had a statistically significant increase in azobenzene modification upon addition of the BDI (p < 0.01).

Figure 5.

Map of PTP1B tyrosine and histidine residues modified by the BDI, with all present tyrosine and histidine residues are highlighted in blue. Modified histidine residues are in red, modified tyrosine residues are in orange. Statistically significant modifications identified by Progenesis have an additional yellow highlight (PDB: 5K9V).

Curious as to why these residues were modified over the others, we looked to the sequence of PTP1B. Given recent reports of BDI derivatives modifying highly reactive tyrosine residues,¹⁷ we hypothesized that the unique microenvironment of these residues renders them prone to reactivity. Starting with the most statistically significant residues, H54 contributes to a magnesium binding site for crystallization,⁴⁹ and is also implicated in celastrol binding, a known inhibitor of PTP1B.⁵⁰ H60 has been shown to be significant in both competitive and non-competitive inhibitor binding via molecular dynamic studies.⁵¹ The tyrosine modified at Y66 is a known hot-spot for protein interactions, and is one of three sites of phosphorylation controlled by insulin levels, crucial for downstream signaling events.^52–54 For the other modifications seen in our MS/MS spectrum analysis, H94 is implicated in the same molecular dynamics inhibitor binding studies as H60, and is also a known hot-spot for protein interactions.⁵⁵ The other tyrosine modified, Y176, is part of the WPD loop and has proven to be integral to WPD loop function. Studies have shown that a Y176A mutation has significantly reduced catalytic activity.⁵⁶ Finally, H214 is an active site histidine residue in the P-loop that been shown to be essential for catalytic activity.^49,57,58 Its important to note that H214 is in close proximity to two other histidine residues, H175 and H173, and a tyrosine residue, Y124, all three of which were not modified. Interestingly, H175 and H173 are not integral to enzyme activity,⁵⁹ and Y124 has been shown to make important interactions near the active site, but is not directly involved in catalysis.⁶⁰ Given the importance of the modified residues, we reasoned that any or all of the observed modifications could lead to inhibition. Taken together, these findings point to a more nuanced reactivity and selectivity of 1 than what was originally expected.

CONCLUSIONS

We presented a novel delivery mechanism of the benzene diazonium ion, a well-established bioconjugation agent for profiling surface tyrosine and histidine residues. The novel protected triazabutadiene technology used in this study proved to be beneficial in delivering the BDI intracellularly. This represents the first report of triazabutadienes being utilized for in cellulo experiments, and the first report of an aryl diazonium ion being targeted intracellularly.

Importantly, these findings provide a platform for a wide range of future studies. The ability to selectively bring reactive electrophiles to specific biochemical environments enables a range of new experiments wherein the accessibility and reactivity of various residues can be interrogated in intact biological systems. This claim is strengthened by the selectivity observed in our mass spectrometry experiments, providing support for a model where the resulting BDI from 2, or variants thereof, could be used as a small molecule covalent probe for Activity-Based Protein Profiling (ABPP).⁶¹ Functional sidechain profiling has been done with cysteine, lysine, and tyrosine, additional methods need to be explored in order to profile functional histidine residues.^62–64 Recent reports of alkyne-containing diazonium ions being used for proteomic profiling of cell lysate underscore the potential for aryl diazonium ions being used in this way.¹⁷ Future investigations into the global cellular effects of treatment with the BDI using pull-down probes and whole-cell proteomics are underway.

Future studies to holistically explore cellular targets that the BDI is modifying are planned, be they protein or DNA. It is plausible too that the BDI induces additional effects to extracellular proteins that also result in observed increase in global tyrosine phosphorylation, and we hope to investigate those phenomena more in the future. While the BDI is a well-established bioconjugation probe, there is still much to be learned regarding selectively so that it that may be leveraged to uncover functional information of proteins. Future studies in this area include modifying the BDI to optimize selectivity for reactive histidine residues. This study provides promise for further development of BDIs in activity-based protein profiling and beyond.

ABBREVIATIONS

BDI

Benzene Diazonium Ion

TLC

thin layer chromatography