|Home | About | Journals | Submit | Contact Us | Français|
tert-Butyl nitrite was identified as a safe and chemoselective nitrating agent that provides preferentially mononitro derivatives of phenolic substrates in the presence of potentially competitive functional groups. On the basis of our control experiments, we propose that the reaction proceeds through the formation of O-nitrosyl intermediates prior to C-nitration via homolysis and oxidation. The reported nitration method is compatible with tyrosine-containing peptides on solid support in the synthesis of fluorogenic substrates for characterization of proteases.
The technique of Fluorescence Resonance Energy Transfer (FRET) is currently a common method for monitoring structural, functional, or aggregation changes in evaluated biomolecules.1 Real-time monitoring of hydrolase activity is a particularly useful application of FRET, because the substrate can be chemically modified at sites remote to the scissile bond, minimizing probe interferences with substrate recognition. The 2-aminobenzoic acid (2-Abz) fluorophore and 3-nitrotyrosine quencher pair has been utilized extensively in determining proteolytic activities of endoproteases.2 As amino acids, these agents represent a convenient fluorophore-quencher combination that can be readily incorporated into a conventional solid phase peptide synthesis protocol. As a part of a program in evolution of site-specific proteases as therapeutic agents, we are interested in developing a flexible and practical platform for kinetic and thermodynamic characterization of evolved enzymes. Thus, we envisioned a unified synthetic strategy for accessing both internally quenched substrates, useful for real-time kinetic analysis of active enzymes, as well as unquenched variants, suitable for fluorescence quenching and anisotropy studies with inactive mutants. We were very keen, therefore, to identify a reagent that could perform a nitration reaction on solid phase-immobilized peptides containing tyrosines with high efficiency and selectivity. Unfortunately, the existing nitrating agents, such as mixed acids, superacids, and metal nitrates, are not suitable for polymer-supported substrates because of their polymer-reactive or heterogeneous nature.3,4 In addition, many of the existing reagents can lead to over-nitration of phenols4a,g,h and other unwanted reactions with a variety of functional groups found in peptides5. Thus, development of a methodology enabling chemoselective nitration of polymer-immobilized tyrosine would be valued. We disclose in this communication a novel nitration procedure using tert-butyl nitrite, a reagent that displays exquisite chemoselectivity for phenols, yields only tert-butanol as a byproduct, and is relatively safe, volatile, and soluble in a variety of organic solvents.
In our attempt to perform diazotization of tyrosine with t-BuONO,6 we observed the formation of unexpected yellow products. To explore this finding further, we subjected Boc-Tyr-OH (1a) to the nitrite treatment in dichloromethane at room temperature (Scheme 1). After 3 hours of agitation, we detected a complete consumption of 1a and quantitative (>95%) formation of Boc-Tyr(3-NO2)-OH (2a) and corresponding N-nitroso derivative 3a (2a:3a 57:43). While there have been reports of C-nitro products emerging from the exposure of electron-rich aromatic compounds to inorganic7 or organic8 nitrites, no mechanistic or scope studies have been carried out. In 1H NMR spectum of the crude reaction mixure we also identified a trace ammount of bisnitrated product 4a (<1%), and we confirmed that both 3a and 4a can arise from 2a by re-exposing the latter to a larger excess of t-BuONO. The presence of a nitro rather than a nitrosyl group in 2a was corraborated by mass spectrometry, 1H NMR,9 and UV-Vis spectroscopy (λmax ≈ 430 nm). Although N-nitrosyl compound 3a rapidly decomposed upon isolation to form 4a along with several other compounds, we confirmed its structure by comparison of 1H NMR resonances with those of the more stable ester variant 2b, prepared from Boc-Tyr-OMe (1b) along with 3b and 4b (2b:3b:4b 47:32:21, >95%) under identical conditions (Scheme 1). On the basis of these results, we speculated that a careful optimization of reaction conditions could provide a selective mononitration procedure.
To identify a more practical nitration procedure, we proceeded to screen a variety of solvents (Table 1). While protic solvents appeared to prevent nitration altogether (entry 1), use of chloroform noticeably improved the chemoselectivity of this reaction (entry 2). Upon prolonged exposure, however, undesired byproducts 3a and 4a became dominant (entry 3), limiting the utility of this solvent. In DMSO, the reaction was noticeably slower than in the halogenated solvents, but provided exclusively the desired product 2a (entries 4 and 5). The reaction proceeded in DMF at a similar rate as in chloroform (entry 6), but displayed higher chemoselectivity for mononitration, since byproducts 3a and 4a emerged only upon extended incubation with t-BuONO (entry 7). The use of acetonitrile, on the other hand, severely compromised chemoselectivity of the reaction with byproducts emerging early in the reaction course (entry 8) and accumulating significantly after 16 hours (entry 9). While the use of diethyl ether led to a relatively slow nitration progress (entry 10), a temperature increase led to nearly quantitative conversion to 2a in 3 hours (entry 11). Finally, by employing THF we obtained a nitration procedure that is both highly selective and reasonably fast. Within one hour, reaction conversion reached 93% (entry 12), and became nearly quantitative after 3 hours (entry 13). While over-nitration product 4a started to form after a 6-hour incubation (entry 14), no detectable N-nitrosylated product emerged even after 16 hours (entry 15). These results demonstrate that the rates of neither the desired C-nitration nor the undesired N-nitrosylation and bisnitration reactions appear to correlate with solvent polarities. Over-nitration has occurred in both polar and nonpolar solvents, albeit at significantly distinct rates: faster in dichloromethane (ε = 8.910) and acetonitrile (ε = 36.6), but slower in DMF (ε = 38.3) and THF (ε = 7.5). Widely varying ratios of the byproducts (cf. entries 3, 7, 8, and 15) suggest distinct mechanistic pathways responsible for their formation. The preference for N-nitrosylation in acetonitrile could be rationalized by the formation of a solvent-nitrosonium adduct,11 capable of the subsequent N-nitrosylation of the carbamate. No byproducts were observed in the aprotic solvents at the extremes of polarity: DMSO (ε = 47.2) and diethyl ether (ε = 4.3). Unlike these solvents, however, THF maintained a robust nitration rate, providing a procedure that was fast, selective, and more suitable for solid phase synthesis, considering favorable swelling properties of crosslinked polystyrene resins in THF.12
To establish the scope of this reaction, we exposed a series of phenols to t-BuONO in THF. We found that this procedure was effective at converting electron-rich compounds in a variety of steric contexts into C-nitro derivatives (Table 2). When both ortho and para positions are available, a mixture of mononitrated regioisomers was obtained (entry 1). The resistance to over-nitration in THF was also confirmed by the lack of reactivity of both o- and p-nitrophenols toward t-BuONO (1.5 equiv, 2 hrs). As expected, ortho- and para-substituted phenols (entries 2 and 3) provided exclusively the corresponding para- and ortho-nitro derivatives, respectively. Notably, single regiochemical outcomes with 4-methoxyphenol and 2,6-dimethoxyphenol as substrates (entries 4 and 5), highlighted a dominant role of hydroxyl groups over alkoxy substituents in directing the reaction regiochemistry.
Next, we determined the extent of chemoselectivity displayed by t-BuONO with respect to functional groups present in peptides. Gratifyingly, aromatic ethers (anisole, Ac-Tyr(Bn)-OMe) and N-formyl indole (Ac-Trp(CHO)-OMe) were unreactive toward the organic nitrite (3 equiv t-BuONO, THF, 25 °C, 24 hrs) confirming the strict requirement for an aromatic hydroxyl group in this reaction. In addition, no nitrosylation of primary or secondary amides was detected upon prolonged incubation (> 6 hrs) of Ac-Ala-OMe and Ac-Gln-OMe. Of all functionalities tested, only the sulfide-containing compounds Boc-Cys(Tr)-OH and Boc-Met-OH yielded complex mixtures, as seen in earlier diazotization attempts of related compounds.13
To unravel the role that the phenolic hydroxyl group plays in this nitration, we subjected 2,4,6-tri-tert-butylphenol 5 to a large excess of t-BuONO (Scheme 2, Part A). Within 10 min, we observed complete consumption of 5 and concomitant formation of a new compound that we have assigned as O-nitroso derivative 6 on the basis of 1H NMR and FTIR data. Furthermore, in t-BuOH as a solvent, 6 reverted to 5 with concomintant formation of t-BuONO, confirming both the reversibility of the O-nitrosylation process and the chemical nature of 6 as aromatic nitrite. Upon storing at room temperature in CDCl3, 6 slowly transformed into a complex mixture of C-nitro derivatives14 through apparent intermediacy of a phenoxyl radical of 5, detected by its characteristic blue color15. The isolation of the O-nitroso species, the lack of reactivity in protic solvents, and the reaction kinetics — first order with respect to both phenol and t-BuONO (see Supporting Information) — allowed us to propose that the reaction is initiated by a rate-limiting nitrosyl exchange between the alkyl nitrite and aromatic hydroxyl groups (Scheme 2, Part B), as shown previously for aliphatic alcohols.16 The aromatic O-nitroso derivatives undergo subsequent thermal homolysis to furnish resonance-stablized phenoxyl radicals and nitric oxide.17 Our attempts to detect any C-nitroso intermediates failed, and we, therefore, propose that the final steps of this reaction involve rapid oxidation of nitric oxide with molecular oxygen,18 prior to coupling of the resultant nitrogen dioxide with phenoxyl radicals.19 A crossover experiment with 6 as a nitrating reagent furnished the mononitro product (4-methoxy-2-nitrophenol) with 4-methoxy-phenol as a substrate (see Supporting Information), confirming the intermediacy of aromatic nitrites in this process. Use of anisole in this experiment, however, left the aromatic ether unmodified, establishing that the formation of phenoxyl radicals is the prerequisite for the observed C-nitration. Participation of phenoxyl radicals has been postulated for nitration reactions mediated by tetranitromethane5a and peroxynitrite20 in polar protic media, on the basis of isolation of biaryl ethers and biphenols as dominant byproducts. We do not observe these products, presumably due to the improved solvation of phenoxyl radicals in the aprotic solvents, making the bimolecular cross-coupling less likely.
Having developed an efficient nitration procedure for solution phase transformations, we sought to implement this methodology on polymer-supported phenolic substrates. We generated test dipeptide 7, containing both protected and unprotected tyrosine residues on Merrifield resin (Scheme 3, Part A) to investigate the ability of the reagent to target selectively the phenol functional group in the presence of carbamate, secondary amide and aromatic ether functionalities. The sole product isolated after the treatment of 7 with t-BuONO in THF and subsequent hydrolysis was Tyr(3-NO2)-containing derivative 8. Finally, we accomplished efficient debenzylation of 8 in the presence of a nitro group under transfer hydrogenation conditions to provide Boc-Tyr(3-NO2)-Tyr-OH (9).
Our interest in Tobacco Etch Virus protease (TEV-Pr) as a scaffold for the evolution of enzymes with novel properties prompted us to explore synthetic approches toward fluorogenic substrates of TEV-Pr. In a search for the shortest TEV-Pr substrate, a tripeptide truncate (FQG) of the optimal linear epitope ENLYFQG,21 flanked by the C-terminal Asp-Tyr dipeptide and N-terminal 6-aminohexanoic acid spacer-fluorophore (Ahx-Abz) unit (10), was produced as a test compound for the solid phase nitration procedure (Scheme 3, Part B). Reaction of 10 with t-BuONO, followed by hydrolysis, yielded a selectively mononitrated peptide furnished with a 3-nitrotyrosine chromophore in > 90% purity, as detected by reverse phase HPLC, spectrophotometric analysis, and mass spectrometry (see Supporting Information). While the truncated substrate was not recognized by TEV-Pr, this stringently quenched peptide (F0/F∞ ≈ 0.9%) was processed effectively by chymotrypsin, as detected in real-time through the release of a fluorescent anthranilamide-containing fragment (see Supporting Information).
In conclusion, we have presented here a novel procedure for selective and efficient nitration of phenols in aprotic media, which is compatible with polymer-supported substrates. We expect that the potential of this methodology to provide access to both internally quenched and unquenched peptides will greatly facilitate characterization of novel proteolytic activities. Finally, the facile and selective introduction of a nitro group described herein may also initiate other synthetic opportunities, such as late-stage chromophore installation and formation of photolabile groups.
This work was funded by NIH grant 1R21AG031437-01.