Enantioselective catalytic synthesis of α-aryl-α-SCF₃-β^2,2-amino acids

Abstract

We herein report a novel entry towards chiral α-SCF₃-β^2,2-amino acids by carrying out the ammonium salt-catalyzed α-trifluoromethylthiolation of isoxazolidin-5-ones. This approach allowed for high enantioselectivities and high yields and the obtained heterocycles proved to be versatile platforms to access other targets of potential interest.

While nature employs only twenty-two genetically encoded proteinogenic amino acids (AAs) as vital building blocks for life, a plethora of chemically synthesized unnatural AAs exists. In principle, the later offer infinite tools to expand the genetic code and to investigate protein’s structure and function. Of particular importance, fluorinated amino acids (FAAs) fall within the realm of specific investigations related to structural and functional learning of peptidic compounds.¹ The introduction of a single fluorine atom or a CF₃ motif are common approaches to obtain FAAs,² whereas the supply of alternative F-containing motifs remains much less-explored (Scheme 1A). The incorporation of the highly lipophilic trifluoromethylthio group CF₃S into amino acids (and subsequent peptides) should be beneficial for the elucidation of structure and function, as well as for pharmaceutical developments. Indeed, as analogues of the sulfur-containing methionine and cysteine, SCF₃-derivatives should help to escape the hydrophobic core where thiol residues are usually buried while ensuring protection against oxidation. Furthermore, thanks to the power of ¹⁹F NMR techniques and some recently developed ¹⁸F-radio-labeling strategies of the CF₃S group, the site-specific installation of trifluoromethylthio amino acids (SCF₃-AAs) to probe local events in peptides is a powerful tool to investigate structure and function in a complementary fashion.³

Scheme 1.

(A) Overview of FAAs and SCF₃-AAs; (B) targeted asymmetric α-SCF₃-β^2,2-AA syntheses starting from α-aryl-isoxazolidin-5-ones 1.

The common strategy to access SCF₃-AAs, relies on the trifluoromethylation of thiols⁴ or disulfides.⁵ As for the direct introduction of the whole CF₃S motif in AAs, rare examples include regioselective trifluoromethylthiolation of unactivated C–H bonds of amino esters,⁶ electrophilic α-trifluoromethylthiolation of glycine Schiff bases,^7a aminotrifluoromethylthiolation of α,β-unsaturated carbonyl compounds^7b and nucleophilic trifluoromethylthiolation of a variety of cyclic sulfamidates based on a stereospecific S_N2-type process (i.e. this strategy developed by one of us gives access to an enantiopure l-threonine derivative where the OH-group is replaced for a CF₃S-group).⁸ However, to the best of our knowledge, no enantioselective approach for the synthesis of chiral SCF₃-AAs was reported until now.⁹

The lack of enantioselective catalytic procedures to access novel SCF₃-AAs led us to study the construction of β^2,2-AAs featuring the CF₃S motif at the tetrasubstituted non-racemizable stereocenter.¹⁰ β-AA have attracted considerable attention over the last years,¹¹ and more recently the easily accessible α-substituted isoxazolidin-5-ones 1 ¹² (as masked β²-AAs) have gained significant interest as precursors for β^2,2-AA derivatives.^9,13–15 Especially, the use of asymmetric phase-transfer catalysis (PTC), by means of chiral quaternary ammonium salt ion pairing catalysts,¹⁶ allowed highly enantioselective α-functionalization reactions of 1 towards α,α-disubstituted isoxazolidin-5-ones 2, albeit with a limited number of electrophiles so far (compounds 1 and 2 in Scheme 1B show the specific substitution patterns we wish to investigate herein but several other derivatives have been reported recently). Importantly, this masked β^2,2-AA derivatives can then undergo facile reductive ring-opening reactions to give β^2,2-AA derivatives^14,15 or may be directly incorporated into peptides by means of the so-called KAHA ligation technique.^14d,15a,17

Based on our recent progress in the utilization of compounds 1 to access masked chiral β^2,2-AAs 2 under chiral phase-transfer catalysis,¹⁴ we now became interested in introducing an asymmetric protocol to access novel α-SCF₃-α-aryl-substituted targets 2 by carrying out the electrophilic trifluoromethylthiolation of α-aryl-1 under chiral ammonium salt catalysis (Scheme 1B). The hereby obtained masked SCF₃-β^2,2-AAs 2 represent a unique entry to SCF₃-β^2,2-AAs, dipeptides or the corresponding SO₂CF₃-analogs.

We started our investigations by carrying out the α-trifluoromethylthiolation of the phenyl-substituted 1a with the established electrophilic CF₃S-transfer reagents I–IV ¹⁸ using Maruoka’s catalysts A ¹⁹ (Table 1 gives an overview of the most significant results of a broad and systematic screening). It should be noted that other chiral ammonium salt catalysts were tried as well, but in analogy to our recent observations,¹⁴ none of these quaternary ammonium salts led to useful enantioselectivities.²⁰

Table 1. Identification of the best-suited SCF₃-transfer reagent, catalyst and conditions for the synthesis of 2a^a.

Entry	Cat. (mol%)	[CF3S]+ (eq.)	Solv.	Base (eq.)	Yieldb [%]	e.r.c (S : R)d
1	A2 (5)	I (3)	Et2O	Cs2CO3 (2)	80	82 : 18
2	A2 (5)	II (3)	Et2O	Cs2CO3 (1.1)	70	75 : 25
3	A2 (5)	III (3)	Et2O	Cs2CO3 (1.1)	0	—
4	A2 (5)	IV (3)	Et2O	Cs2CO3 (1.1)	0	—
5	A1 (5)	I (1.5)	Et2O	Cs2CO3 (1.1)	77	90 : 10
6	A1 (5)	I (1.2)	Et2O	K2CO3 (0.2)	89	89 : 11
7	A1 (5)	I (1.2)	iPr2O	K2CO3 (0.2)	85	87 : 13
8	A1 (5)	I (1.2)	THF	K2CO3 (0.2)	66	76 : 24
9	A1 (5)	I (1.2)	CH2Cl2	K2CO3 (0.2)	53	75 : 25
10	A1 (5)	I (1.2)	PhCH3	K2CO3 (0.2)	56	84 : 16
11	A1 (5)	I (1.2)	Et2O	KOAc (0.2)	25	90 : 10
12	A1 (5)	I (1.2)	Et2O	NaOPh (0.2)	71	89 : 11
13	A1 (5)	I (1.2)	Et2O	K2HPO4 (0.2)	88	90 : 10
14	A1 (5)	I (1.05)	Et2Oe	K2HPO4 (0.2)	82	91 : 9
15f	A1 (5)	I (1.05)	Et2Oe	K2HPO4 (0.2)	90	93 : 7 (98 : 2)g
16f	A1 (5)	I (1.05)h	Et2Oe	K2HPO4 (0.2)	90	95 : 5
17f,i	A1 (3)	I (1.05)	Et2Oe	K2HPO4 (0.2)	90	93 : 7

^aAll reactions were run for 20 h at room temperature using 0.1 mmol 1a in the indicated solvent (0.1 M with respect to 1a) with the given reagents and catalysts unless otherwise stated.

^bIsolated yields.

^cDetermined by HPLC using a chiral stationary phase.

^dAbsolute configuration was determined by X-ray diffraction analysis of single crystals of enantioenriched (S)-2a.²¹

^e0.05 M with respect to 1a.

^fRun at –20 °C.

^gAfter recrystallization.

^hAdded with a syringe pump over 5 h.

ⁱ1 mmol scale.

First experiments with catalyst A2 and the succinimide-based reagent I allowed for a promising initial enantiomeric ratio of 82 : 18 already (entry 1). In addition, reagent I performs slightly better than the phthalimide-based II, while neither the saccharin-derived III nor the cumyl O-SCF₃ reagent IV did allow for any product formation at all (entries 1–4). Using the catalyst A1, substituted by the 3,4,5-trifluorophenyl pendant at 3,3′-positions, a slight increase in enantioselectivity could be achieved, combined with lowering the excess of reagent I (entry 5). Even more interestingly, the reaction allowed the use of catalytic amounts of base, without affecting the enantioselectivity (entry 6). The evaluation of different solvents showed that ethers in general furnished higher selectivities than aromatic or chlorinated solvents (entries 6–10). As Et₂O performed slightly better than other ethers, the base-optimization was performed in Et₂O subsequently (entries 11–13). Noteworthy, the nature of the base did not affect the enantioselectivity significantly, contrary to the yield.

As the base can be used catalytically in these reactions, which is rather similar to our recent observations for asymmetric α-sulfanylation reactions of compounds 1,^14a it can be rationalized that the in situ formed K-succinimide may also serve as a base during these reactions (thus rationalizing why different bases resulted in more or less the same enantioselectivities). Then, we carried out one reaction using a catalytic amount of K-phthalimide under the final optimized conditions (vide infra) and the observed selectivity and yield were exactly the same, thus substantiating this mechanistic proposal.

As K₂HPO₄ gave somewhat higher yields than the other bases (entry 13), we carried out the final optimization with this base in Et₂O. It turned out that lowering the reaction temperature to −20 °C, with a slightly higher dilution and slow addition of I provided enantioselectivities up to 95 : 5 (entries 15 and 16). To our delight, product 2a can be further enantioenriched by recrystallization and can also easily be accessed on 1 mmol scale with 3 mol% of catalyst A1 in high yield and enantioselectivity (entry 17) (lower catalyst loadings then lead to significantly reduced conversions). The absolute configuration of product 2a was unambiguously assigned by single crystal X-ray analysis²¹ and the sense of enantioinduction is in line with our previous observations using Maruoka catalysts A with pronucleophiles 1.¹⁴

With high yielding and highly enantioselective catalytic conditions at hand, we next investigated the application scope of this approach for a variety of variously substituted α-aryl-derivatives 1 (Scheme 2). For ease of experimental operation, we carried out the trifluoromethylthiolation reactions under the conditions given in entry 15 (Table 1) avoiding the electrophile addition via syringe pump. As outlined in Scheme 2, a variety of novel masked α-SCF₃-α-aryl-β^2,2-AAs 2a–2k were readily obtained in high yields and with satisfying enantiomeric ratios, ranging from 91 : 9 to 95 : 5 e.r., regardless of the nature of the aryl moiety.

Scheme 2.

Application scope (all reactions were carried out on 0.1 mmol scale under the conditions given in Table 1, entry 15).

To demonstrate the versatility of the herein accessed heterocyclic platform 2, we also carried out the transformations of compound 2a as depicted in Scheme 3. The N–O-bond could be easily cleaved under Pd-catalyzed hydrogenation conditions to afford the free carboxylic acid 4 in literally quantitative yield.^12,14 In addition, the CF₃S-group could be oxidized to the corresponding sulfone (giving product 5) straightforwardly with NaIO₄ under RuCl₃-catalysis.²² Furthermore, compound 2a can directly be engaged into amide bond forming reactions by carrying out a nucleophilic ring opening reaction with benzylic amines to give compounds 6. On the other hand, after a facile N-Boc deprotection of 2a, the KAHA ligation¹⁷ with reagent 8 occurred to give access to the corresponding dipeptide 7 under previously described aqueous-conditions.^14d,15a All these performed transformations furnished high yields under robust reaction conditions, thus substantiating the synthetic versatility of our SCF₃-containing heterocyclic platform 2.

Scheme 3.

Further transformations of compound 2a.

Conclusions

We have shown that α-aryl-isoxazolidin-5-ones 1 can undergo α-trifluoromethylthiolation reactions with high enantiomeric excesses and excellent yields under asymmetric phase-transfer catalysis conditions by using the commercially available Maruoka catalyst A1.⁹ These α-heterofunctionalization reactions perform under operationally simple conditions and could easily be carried out on 1 mmol scale as well. In addition, the hereby accessed masked α-SCF₃-β^2,2-AAs 2 can be utilized to obtain unprotected α-SCF₃-β^2,2-AAs, amides, dipeptides and the corresponding SO₂CF₃-analogs either.

Supplementary Material

† Electronic supplementary information (ESI) available: Full experimental details and analytical data. CCDC 1914375. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c9ob02666e