Studies towards a biomimetic synthesis of α -cyclopiazonic acid: synthesis of 5-substituted isoxazole-4-carboxylic esters

An efficient, high yielding synthesis of ethyl 5-hydroxymethyl-3-methylisoxazole-4-carboxylate has been developed, based on a procedure by Gelin which involves reaction of ethyl acetoacetate with chloroacetyl chloride followed by treatment with hydroxylamine hydrochloride. The product of this reaction was then converted into the bromide and reacted with tetrahydrothiophene to give sulfonium salts in up to 71% overall yield (from ethyl acetoacetate). The synthesis is suitable for use with a chiral sulfide and for large-scale use. The synthesis of ethyl 5-formyl-3-methyl-4-isoxazolecarboxylate and the corresponding tosylhydrazone are also reported. These isoxazoles are starting materials for a proposed convergent, biomimetic synthesis of α -cyclopiazonic acid.


Introduction
α-Cyclopiazonic acid (1) is a mycotoxic metabolite produced by Penicilum cyclopium Westling. 1,24][5] α-Cyclopiazonic acid is currently synthesized using fermentation, but a chemical synthesis would allow variation of the structure in a controlled way.The biological activity and structural complexity of 1 make it an attractive and challenging target for total synthesis.][10] All three make use of pyrrolidine-2carboxylate 2 as a late-stage intermediate.The most recent synthesis by Hashkins and Knight is the most efficient and uses a high-yielding carbocationic cascade double cyclization to synthesize two of the non-indole rings in its key step.No enantioselective syntheses have been reported to date.
We have considered a biomimetic approach to 1.It has been shown that tryptophan and βcyclopiazonic acid are biosynthetic precursors of 1 (Scheme 1). 11,12Steyn and his co-workers have established that the 3-pro-S-hydrogen atom from tryptophan is lost in the cyclization step and confirmed that the formation of the new C-C bond occurs from the side opposite to the proton removal. 13,14However, the mechanism of this oxidative cyclization is not fully understood.
As indicated above, we have engaged in a program to realize a biomimetic enantioselective synthesis of α-cyclopiazonic acid.9][20] The tetramic acid moiety should be accessible by hydrogenolysis of the isoxazole compound 7. [21][22][23][24] The desired aziridine can be disconnected to two combinations of imine and a sulfonium ylide, either 9 and 10 or 11 and 12. Herein we describe our first steps towards the synthesis of α-cyclopiazonic acid -the synthesis of 10 and 12.Although literature precedent suggested that these seemingly 'routine' heterocycles should be accessible in a straightforward manner, we encountered considerable difficulty in making them and so now report a simple, high yielding route to these important intermediates. 22unctionalized isoxazoles 10/12 can be disconnected in a number of ways as indicated in Scheme 3. We initially focused on disconnection a since bromination of isoxazole 13 had been reported. 21We also anticipated that isoxazole 13 might be a common precursor for the synthesis of 10 and 12. Scheme 3. Retrosynthetic analysis of isoxazoles.
In addition, it was found that transformation of either 13 or bromide 14 into aldehyde 16 was more troublesome than expected.Direct oxidation with a variety of oxidizing agents failed (Scheme 5), 29,30 despite literature precedent for the oxidation of ethyl 3-aryl-5bromomethylisoxazole-4-carboxylates. 28Scheme 5. Attempts at direct oxidation of the bromide and the dimethyl isoxazole.
Finally, the synthesis of alcohol 21 was achieved by modifying a procedure of Gelin and coworkers, [34][35][36] to obtain the furanone 22, and then treating the crude product with hydroxylamine hydrochloride (disconnection c, Scheme 7).We were able to improve significantly on the literature yield of 22 (50%). 34We found that the low yields were due to incomplete reaction of 24b.Therefore, after addition of ethyl chloroformate at -10 °C, the reaction was allowed to warm to room temperature and was stirred for a further 2 hours.This allowed us to obtain the alcohol 21 in 79% yield over two steps.

Scheme 7. Synthesis of isoxazole 21 via a furanone.
The isoxazole sulfonium salt 27 was our next target.In many cases, sulfonium salts can be synthesized directly from the corresponding alcohols by treatment of a suitable mineral acid (e.g., HBF 4 or HPF 6 ). 37,38A variety of conditions was tested but the isoxazole 21 was unreactive, probably because it is rather electron-deficient (this reaction works best with electron-rich aromatics).As such, alcohol 21 was first converted into bromide 14.By using PBr 3 and DMF in toluene this transformation occurred in quantitative yield. 39The use of DMF was essential to the success of this reaction; the use of PBr 3 in the absence of DMF, or indeed other brominating agents, did not lead to any bromination.Scheme 8. Improved synthesis of bromide 14.Reagents and conditions: PBr 3 (1.6 eq.), DMF (1.6 eq.), toluene, reflux, quantitative.The bromide 14 was transformed into the sulfonium salt 25a in 61% yield by treatment with tetrahydrothiophene (THT).The sulfonium bromide has a tendency to revert to starting materials and therefore it was necessary to exchange the counter-ion.After some experimentation (Table 1), it was found that this was best achieved by stirring the bromide in acetone with the appropriate sodium salt. 40This procedure furnished sulfonium salts 25b-d in yields of 55%, 65% and 91%, respectively.Attempts to achieve this salt formation and counter-ion exchange in one pot led to lower yields.The isoxazole hydrazone salt 43 was our next target.Oxidation of alcohol 21 using PCC 35 was sluggish (62% yield after one week).Oxidation using Dess-Martin periodinane (DMP) in dichloromethane gave a similar yield after 15 min, but longer reaction times led to decomposition of the product.After two weeks PCC oxidation gave an 82% yield of the desired aldehyde 16 (Table 2).
The aldehyde 16 was treated with tosylhydrazide to give the tosylhydrazone 26 in 76% yield (Scheme 9).Treatment of 26 with sodium methoxide (generated in situ from Na and MeOH) resulted in the tosylhydrazone sodium salt 27, which underwent decomposition to the diazo compound 28 during work up (even at low temperature).Thus, if our catalytic aziridination methodology was to be employed it would have to be via either the direct addition of the diazo compound or alternatively via the in situ deprotonation of the tosylhydrazone salt using LiHMDS or NaHMDS at low temperature. 15,41We now have a successful route to the functionalized isoxazoles, which are important and useful precursors for synthesis.Furthermore, we have prepared the corresponding tosylhydrazone and sulfonium salts -substrates for the catalytic and stoichiometric aziridination protocols, respectively.Work on synthesis of the appropriate indole moieties is ongoing and we hope to report a completed synthesis of α-cyclopiazonic acid in due course.

Experimental Section
General Procedures.Anhydrous benzene was obtained from Aldrich.Dried CH 2 Cl 2 was obtained from the dry solvent dispenser in the University of Bristol, which is built by Anhydrous Engineering, based on the Grubbs design. 42Acetonitrile was freshly distilled from calcium hydride.All reagents were either used as received from commercial sources or purified using recognized methods.Reactions requiring anhydrous conditions were performed using oven-dried glassware under an atmosphere of dry nitrogen.TLC was performed on aluminum sheets precoated with silica (60 F254).The plates were developed using standard visualizing agents.Flash column chromatography was performed using silica gel (Merck Silica gel 60 F254, 230-400 mesh). 1 H NMR spectra were recorded on Jeol Delta 270 (270 MHz), Delta 400 (400 MHz) or ARKAT USA, Inc.
ecp 400 (400 MHz) spectrometers, using dry CDCl 3 , or CD 3 OD as the solvent.Coupling constants, J, are quoted in Hertz (Hz).Chemical shifts (δ) are quoted in parts per million (ppm) and are relative to tetramethylsilane (δ = 0 ppm) as the internal standard in CDCl 3 or the residual solvent peak in CD 3 OD (δ = 3.42 ppm). 13C NMR spectra were recorded on a Delta 400 (100 MHz) or ecp 400 (100 MHz) spectrometer.Melting points were measured on a Köfler Hot stage Micro Melting point apparatus and are uncorrected.Elemental analyses were performed on a Carlo Erba EA1108 or a Perkin Elmer 2400 CHN elemental analyzer.Samples for infra red spectroscopy were prepared using the neat product and the spectra recorded using a Perkin-Elmer 157G FT-IR spectrometer.Only characteristic absorbencies (ν max ) are reported in cm -1 .Both low-and high-resolution mass spectra (m/z) were recorded with a Fisons/VG Analytic Autospec System, with only molecular ions ([M] + or [M+H] + ) and major peaks being reported with intensities quoted as percentages of the base peak.GCMS retention times (R t ) and low resolution MS were recorded on an Agilent 6890 GC (column: HP190915-433 HP-5MS 5% Phenyl Methyl Siloxane; capillary 30 m × 250 µm × 0.25 µm nominal; carrier gas: helium 1 mL/min (constant flow mode); injector: 250 °C (splitless mode).Mass spec.detector: Agilent MSD 5973 (EI mode); oven: 70 °C (3 min), 15 °C/min (15.3 min), 300 °C (8 min).
Ethyl 2-methyl-4-oxo-4,5-dihydro-3-furancarboxylate (22). 34,43The furanone 22 was prepared following a modified literature procedure. 34,43Ethyl acetoacetate (5.01 g, 38.5 mol) was added to a suspension of magnesium ethoxide (6.04 g, 42.3 mol) in anhydrous benzene (9 mL), the mixture was stirred for one hour at room temperature.Anhydrous acetonitrile (9 mL) was added to the reaction mixture and the flask was cooled to -10 °C, followed by the slow addition of αchloroacetyl chloride (3.32 mL, 38.5 mol), after which the mixture was allowed to warm to room temperature and left to stir for two hours.A dilute solution of sulfuric acid (1 mL acid in 35 mL ice/water) was added, followed by extraction with diethyl ether.The combined organic fractions were dried over MgSO 4 and filtered, the filtrate was then cooled to 0 °C and a solution of triethylamine (5.4 mL) in diethyl ether (3 mL) added.The reaction mixture was left to stir at room temperature overnight.The precipitated triethylamine hydrochloride salt was filtered off and the diethyl ether evaporated under vacuum to yield a yellow gum (6.09 g, 93% crude yield).A small amount was recrystallized from diethyl ether which yielded a white solid for characterization; mp 76.5-77.5 °C (Et 2 O); R f 0.5 (EtOAc/Petroleum ether, 2:3); 1 H NMR (400 MHz, CDCl 3 ) δ H 1.39 (3H, t, J = 6.84,CO 2 CH 2 CH 3 ), 2.45 (3H, s, H 3 CC=C), 4.37 (2H, q, J = 6.84,CO 2 CH 2 Me), 4.91 (2H, s, OCH 2 C=O); 13

Table 1 .
Synthesis of isoxazole sulfonium salts a Reagents stirred neat in one pot; b Reagents stirred in one pot in minimum acetone; c 25a treated with NaBF 4 in minimum acetone.