Synthesis of 1,2,3-triazol-1-yl-methaneboronic acids via click chemistry: an easy access to a new potential scaffold for protease inhibitors

Stereoselective synthesis of previously unreported 1,2,3-triazol-1-yl-methaneboronic acids has been achieved from azidomethaneboronates by Copper-catalyzed Azide-Alkyne Cycloaddition (CuAAC). The proximity of the cycloaddition reaction center to the boronic group is not detrimental for the stability of the sp3-carbon-boron bond nor to the stereoisomeric composition, further expanding the field of application of click chemistry to new boronate substrates and offering a new potential scaffold for protease inhibitors.


Introduction
α-Amidomethaneboronic acid is a recurring core-structure in biologically active and important boron-containing compounds. [1] After the approval of Velcade® (Figure 1a, A) and clarification on safety issue relatively to boron containing compounds, the use of this element in pharmaceutical research has become an attractive "hot" topic. As a result, several boron derivatives are currently in preclinical and clinical stage development. [2] Amongst others boronic acids exhibit excellent properties as competitive and reversible protease inhibitors. Due to their unique structural features, they act as transition state analogs: boron, with its open shell, interacts with nucleophilic active residues and in doing so converts from a neutral trigonal structure to an anionic tetrahedral adduct, which mimics the high-energy intermediate of the amide hydrolysis process. Whereas boron moiety acts as the "warhead" blocking the catalytic site, the α-amido group enhances molecular recognition by mimicking natural substrates. [3] α-Amidomethaneboronate unit is the basic structure of peptidoboronic acids, a class of peptidomimetics largely explored to target different clinically relevant proteases. For example, the anticancer Velcade® (A) is a dipeptidyl boronic acid (Phe-boroLeu) acting as stability to hydrolysis, that allows for instance the use of TLC as reaction monitoring method. First investigation of CuAAC feasibility on the azido intermediate 2 was performed choosing ethyl propiolate as acetylene counterpart, according to the general observation that alpha-carbonyl-are more reactive than alkyl-or aryl-alkynes. [15] Among the wide variety of conditions described in the literature for CuAAC, we selected three. In two cases the copper(I) catalyst was directly added in the presence of a ligand (CuI, DIPEA, THF, or CuI, lutidine, CHCl 3 ); [16] in the third case the catalytically active metal was generated in situ by reduction of copper sulfate (CuSO 4 , sodium ascorbate, tert-butanol, H 2 O). [8] Each experiment was performed at room temperature for 6 hours with a molar ratio of 2 : ethyl propiolate : catalyst 1 : 1.5 : 0.1. The crude product was analyzed by 1 H-NMR and LC-MS and the formation of the expected and previously unreported product 3a in almost complete conversion was observed in all of the three experiments, confirming the robustness of CuAAC. Nevertheless, when the CuI catalyst was adopted, the NMR spectra revealed the presence of proto-deboronation by-products (5-20%), and these were more pronounced when the more basic DIPEA rather than lutidine was used as ligand. However DIPEA could be easily removed from the crude under reduced pressure, while lutidine was not. A superior performance in terms of purity of the recovered material and absence of deboronation byproducts was observed for the aqueous conditions, which were therefore applied to cycloadditions of 2 with several other alkynes, selected among the many available on the market, including carbonyl, aromatic and aliphatic alkynes. In the optimized procedure, the azide 2 and an excess of the alkyne (1.5 equiv.) were dissolved in a 1:1 mixture of tertbutanol and water, together with copper sulfate (0.05 equiv) reduced in situ by sodium ascorbate (0.2 equiv.). Cyclizations were carried out at room temperature and followed by TLC until disappearance of the starting azidomethaneboronate 2: complete conversions were reached in two hours with propiolic acid and ethyl propiolate (Table 1, entries 1-2), while longer reaction times (up to sixteen hours) were required for alky-and aryl-alkynes (Table 1, entries 3-5). The expected 1,4-disubstituted triazoles were easily isolated by extraction and removal of the residual alkyne under reduced pressure, affording 3a-e in good to excellent yields (85-99%) as highly pure material. Cyclization was confirmed by a singlet downfield in the aromatic region in the 1 H NMR spectra and the expected 1,4-regioselectivity was supported by bidimensional spectroscopy (particularly the 3 J C,H correlation between protons on the boron-bearing carbon atom and the unsubstituted carbon of the triazole ring). Final deprotection of (+)-pinanediol was accomplished by transesterification with phenylboronic acid in a biphasic system acetonitrile/n-hexane, [17] allowing to obtain final boronic acids 4ae, purified by crystallization from acetonitrile (80-100%).
The successful synthesis of these 1,2,3-triazol-1-yl-methaneboronic acids prompted us to expand our project toward chiral compounds introducing an R 2 -substituent (see Figure 1b). To obtain an homochiral series with natural amino acids a stereoselective synthesis was required. At first we focused on triazolyl analogs of acyl-boroLeu bearing an isobutyl as R 2 group (Scheme 2), a structure that is also part of Velcade® (A).
The same procedure was replicated for the synthesis of boroPhe analogs 18a-e and 19a-e (Scheme 3), bearing as R 2 a benzyl or its meta-carboxy derivative, this latter being a recurring motif in β-lactamase inhibitors (Figure 1, C).
(+)-Pinanediol boronates 10 and 11 were subjected to two consecutive homologation steps: the first with chloromethyllithium for methylene insertion to 12 and 13, the second with dichloromethyllithium to introduce the halogenated carbon atom (14 and 15). [17] Chlorine substitution with sodium azide in phase transfer conditions afforded 16 and 17. These key azido intermediates were then subjected to click reaction and deprotection to triazolyl analogs of acyl-boroPhe 18a-e and 19a-e (Table 2, entries 6-15).
The results reported in Table 2 indicates that the described procedure is reproducible and highly efficient, affording in all cases the expected triazolylmethaneboronic acids in moderate to good overall yields as pure and stable solids that can be stored for months at +4 °C. The rate of CuAAC is not affected by the structure of azidomethaneboronates, but only by the electronic density of the alkyne partner: for a given alkyne, reaction times in fact are consistent for both primary (intermediate 2) and secondary azides (intermediates 16 and 17). Furthermore, for these latter derivatives cycloaddition reaction proceeds without any change in the diastereoisomeric composition, eventually affording enantiomerically-pure triazolylmethaneboronic acids.

Conclusions
A synthetic procedure for enantiomerically-pure 1,2,3-triazol-1-yl-methaneboronic acids has been developed. This new scaffold can be obtained through CuAAC between stereoisomerically pure 1-azidoalkylboronates and terminal acetylenes, catalyzed by copper sulfate reduced in situ to Cu(I) by sodium ascorbate in tert-butanol/water system. In these conditions, the proximity of reaction center to the boronic group is not detrimental to the stability of the sp 3 carbon-boron bond to copper(I) catalysis, further expanding the functional group compatibility of CuAAC beyond what is already known. Application of powerful click chemistry to boronates enables many analogs to be synthesized quickly.
Given the importance of α-amidomethaneboronic acids as proteases inhibitors, this efficient access to a new bioisosteric scaffold could promote further exploration of boronates as a promising class of biological active compounds.

General methods
All reactions were performed under argon using oven-dried glassware and dry solvents. Dry tetrahydrofuran (THF) was obtained by standard methods and freshly distilled under argon from sodium benzophenone ketyl prior to use. All of the reagents were used as purchased from commercial suppliers without further purification. The -100 °C bath was prepared by addition of liquid nitrogen to a pre-cooled (-78 °C) mixture of 1:1 ethanol/methanol. Preloaded (0.25 mm) glass supported silica gel plates (Kieselgel 60, Merck) were used for TLC analysis, and compounds were visualized by exposure to UV light and by dipping the plates in 1% Ce(SO 4 )·4H 2 O, 2.5% (NH 4 ) 6 Mo 7 O 24 ·4H 2 O in 10% sulfuric acid followed by heating on a hot plate. Melting points were measured in open capillary tubes on a Stuart SMP30 Melting Point apparatus. Optical rotations were determined at +20 °C on a Perkin-Elmer 241 polarimeter and are expressed in 10 −1 deg cm 2 g −1 . 1 H and 13 C NMR spectra were recorded on a Bruker Avance-400 MHz spectrometer. Chemical shifts (δ) are reported in ppm and were calibrated to the residual signals of the deuterated solvent. [20] Multiplicity is given as s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br = broad signal; coupling constants (J) are given in Hz. Two-dimensional NMR techniques (COSY, HMBC, HSQC) were used to aid in the assignment of signals in 1 H and 13 C spectra. Particularly, in the 13 C spectra, the signal of the boron-bearing carbon atom tends to be broadened, often beyond the detection limit; however, its resonance was always unambiguously determined by HSQC. Also the triazole ring carbon signals are often beyond the detection limit; when possible these were determined by HSQC and HMBC. High-resolution mass spectra were recorded on an Agilent Technologies 6520 Accurate-Mass Q-TOF LC/MS; elemental analysis were performed on a Carlo Erba Elemental Analyzer 1110.
The enantiomeric purity of chiral boronic acids was checked by reconversion into their pinanediol esters. In particular final compounds 9a-e, 18a-e and 19a-e were allowed to react with an equimolar amount of (+)-pinanediol in anhydrous THF: the NMR spectra of the crude products displayed the presence of a single diastereoisomer, proving that no epimerization occurred during transesterification.