Amidation Strategy for Final-Step α-Hydroxytropolone Diversification

Abstract

α-Hydroxytropolones (αHTs) are excellent metalloenzyme-inhibiting fragments that have been the basis for the development of potent inhibitors of various therapeutically important enzymes. The following manuscript describes a final-step amidation approach for αHT diversification. The method takes advantage of a scalable, chromatography-free synthesis of a carboxylic acid-appended αHT, and in the present manuscript we describe the synthesis of eight amide-containing αHTs, three of which we envision using as chemical probes. We expect that the general strategy will find widespread usage in both chemical biology and medicinal chemistry studies on αHTs.

α-Hydroxytropolones (αHTs, Scheme 1A) are a class of heavily oxygenated aromatic compounds featuring a cycloheptatrienone ring and two free hydroxyls, with all three exocyclic oxygen atoms in a contiguous array. This unique arrangement of oxygen atoms has been shown to make αHTs potent inhibitors of a variety of therapeutically-relevant dinuclear metalloenzymes (Scheme 1A).¹ Our lab has developed an oxidopyrylium cycloaddition/ring-opening approach to the class of molecules (Scheme 1C)^2,3 that has proven to be useful in a broad range of structure function and optimization-driven medicinal chemistry studies.⁴ One example of this activity is 6 (Scheme 1B), which has potent and selective cellular antiviral activity against hepatitis B,⁵ and which recently demonstrated modest but statistically significant antiviral activity in an FRG Human Liver Chimeric Mouse model. ⁶

Scheme 1. α-Hydroxytropolone overview.

( A) α-Hydroxytropolone (αHT), redrawn in its dianionic tropylium form to illustrate dinuclear metalloenzymes inhibition. (B) A synthetic αHT and its anti-hepatitis B activity and cytotoxicity against the non-infected viral host cell. (C) Overview of oxidopyrylium cycloaddition/ring-opening route to αHTs.

Despite the method’s advantages, one of the major limitations is the harsh ring-opening (3→4) and demethylation (4→5) conditions, which prevents the incorporation of many valuable acid-labile substituents. While a recently described 3-component oxidopyrylium cycloaddition allowed for the incorporation of base-labile groups in place of the methyl ether,⁷ the route suffered from lower yields and does not bypass the acid-mediated ringopening step of the sequence. We recently exploited this 3-component method in a fluorous-phase synthesis of αHTs, and through these studies found that amide-coupling could be performed on the free tropolone prior to cleavage of a fluorous tag.⁸ However, cleavage of the tag also required similarly harsh conditions.

In the current manuscript, we demonstrate that a similar amidation can be performed on a scalable, fully non-protected αHT, thus providing a means to access various substituted amide-containing αHTs in the final step. We demonstrate the practicality of this method to access various tropamides with acid-sensitive groups, including molecular probes for future proteiomics studies.

Carboxylic acid αHT 5a had been previously described,⁹ and in the present manuscript we present a scale-up of the procedure that eliminates chromatography. Methyl ester oxabicycle 3a was synthesized in large-scale quantities via a 1,3-dipolar cycloaddition with oxidopyrylium dimer 7 and an excess of methyl propiolate (2b). After removal of the solvent and propiolate in vacuo, 3a can be immediately subjected to triflic acid-promoted ring-opening conditions. This process provided over 5 g of penultimate intermediate 4a that could be demethylated/hydrolyzed using hydrobromic acid in acetic acid and water to generate the carboxylic acid-containing αHT, 5a. This latter step has been accomplished successfully on a moderately large scale to generate 260 mg of the carboxylic acid 5a, which at the current time is a convenient scale for our studies. Efforts to provide gram scale synthesis of this compound are currently underway.

Our amidation conditions were closely derived from those used previously in our fluorous-phase synthesis,⁸ and was performed as follows: Acid 5a was mixed with 2,6-lutidine and PyBOP¹⁰ in THF under argon for 15 minutes at room temperature in order to activate the carboxylate for nucleophilic attack (5a→8, Scheme 3). The amine was then added to the reaction vessel, at which point the vessel was sealed and heated in a microwave reactor at 85°C for 10 minutes. Following column chromatography on C18-capped silica, the amide products were recovered in moderate to good yields. We believe these short reaction times, in addition to the scalability of the intermediates, should allow for rapid αHT library synthesis moving forward.

Scheme 3.

Amidation of tropolone carboxylate 5a

Amides derived from benzylamine (9c) and piperidine (9a) were synthesized in 60% and 78%, respectively, demonstrating the ability for primary and secondary amide synthesis. The coupling of 4-phenylbenzylamine was successfully completed using a slightly varied approach that overcame co-elution of PyBOP-based byproduct tris(pyrrolidinophosphine) oxide with non-polar substrates on reverse phase silica. HBTU couplings¹¹ feature a mildly water-soluble byproduct, tetramethyl urea, which was easily separable via column chromatography. The solvent was also changed to EtOAc due to HBTU’s improved solubility in it as compared to THF. Using this method we were able to obtain biphenyl amide αHT 9d in 50% yield. A fluorescent substrate, 9h, was also synthesized, which we envision could be useful, for example, to study αHT localization in cells,¹² or in the development of a fluorescence polarization assay for screening ligands against various dinuclear metalloenzymes.¹³

Another benefit to the route was its effectiveness at generating αHTs with acid-sensitive functional groups. For example, N-Boc piperazine-derived amide 9b was synthesized in 53% yield. Notably, the acid-labile Boc group remained intact during the reaction and subsequent column chromatography. It was easily removed in a quantitative manner via microwave irradiation with trifluoroacetic acid in CH₂Cl₂ for 2 minutes at 70°C to yield the trifluoroacetate salt 10 (Scheme 4).

Scheme 4.

Deprotection of αHT 9b

The method also provides access to two αHT probes of interest for proteiomics experiments that have acid-labile appendages that are vital to their efficacy as useful probes. Biotinylated probes such as 9f require the use of a PEGylated spacer in order to allow ternary complex formation between avidan/streptavidid and the protein of interest.¹⁴ The diazirine in probe 9g is used for photochemical, covalent modification of enzyme targets in live cells, while the alkyne in 9g is then used to label the enzyme with a reporter group through copper-catalyzed azide-alkyne cycloaddition.¹⁵ Both of these compounds are unattainable via our previously described αHT synthetic routes due to their incompatibility with the harsh acidic environment of the demethylation.

Using the new route, we were able to obtain these molecules, albeit in <40% yields. A lower yield was also observed with aniline 9e, which was obtained in only 38%. However, we found that allowing the reaction to run at room temperature over several days improved yields substantially in the cases of the aniline (38% to 72%) and the diazirine (35 to 61%). Less significant yield increases were found in the case of the biotinylated αHT 9f, which we suspect may be due to solubility. Running reactions at room temperature is a practical method for parallel library synthesis, which could have strategic value moving forward.

In conclusion, we have disclosed conditions for the chromatography free, scalable synthesis of a carboxylic acid-containing αHT, 5a, and have shown that amidation couplings can be carried out under either thermal or room temperature conditions. This method provides a reliable way to incorporate previously inaccessible acid-sensitive functional groups onto the αHT core, and has led to the synthesis of three αHT chemical probes, 9f-h. The divergent approach is also a promising method for facile tropamide library synthesis, and current efforts are underway to leverage this in a variety of medicinal chemistry studies.

Scheme 2.

Chromatography-free synthesis of 5a

Highlights.

1. 7 step, single chromatography synthesis of amide-containing hydroxytropolones

2. Amidation is the final step of the synthesis, providing rapid diversification

3. Chromatography free, gram scale synthesis of penultimate intermediate

4. Successful synthesis of primary, secondary, and aniline-based tropamides.

5. Synthesis of new and previously inaccessible probes for proteomics.