Enantioselective Synthesis of Spiro[cyclohexane-1,3′-indolin]-2′-ones Containing Multiple Stereocenters via Organocatalytic Michael/Aldol Cascade Reactions

Abstract

Organocatalytic reactions of 3-olefinic oxindoles and pentane-1,5-dial were investigated to provide access to substituted spirocyclohexane oxindoles via Michael/Aldol cascade reactions. Of particular interest, we have examined the stereochemical outcome of electron withdrawing and electron-donating groups on the oxindole ring nitrogen. Interestingly, we have observed that the N-protecting group on the oxindole has critical effect on aldol ring closure leading to ultimate stereochemical outcome of the hydroxyl center. The overall process is quite efficient and afforded products with multiple stereocenters in high yields and excellent enantioselectivities (>99% ee).

Spiro[cyclohexane-1,3′-indolin]-2′-one is an important structural motif for many bioactive natural products as well as for medicinal agents.^1,2 Over the years, various stereoselective synthetic protocols have been developed.^3-5 However, enantioselective and efficient methods for construction of these structural features did not emerge until recently.^6-9 Organocatalytic cascade reactions leading to functionalized cyclohexane rings were earlier developed by Enders et. al.¹⁰ In the context of our design and synthesis of molecular probes, we have incorporated a spiro-indoline structural template as a P2′-ligand in the HIV-1 protease active site.¹¹ For further access to spiroindoline scaffolds, we have investigated the feasibility of organocatalytic Michael/aldol cascade reactions involving 3-olefinic oxindole and pentane-1,5-dialdehyde.^12-15 Recently, Wang and co-workers have reported a related work and their report prompted us to disclose our investigation in this area.¹⁶ Herein, we report (R)-diphenylprolinol silyl ether-catalyzed tandem Michael/aldol reaction leading to the synthesis of a variety of spiro[cyclohexane-1,3′-indolin]-2′-one derivatives with four/five contiguous chiral centers in excellent yield and optical purity. Interestingly, protecting groups on the indolin-2-one nitrogen played important roles in the stereochemical outcome of the final aldol ring closure. The electron-withdrawing N-protecting group in the presence of (R)-catalyst provided spiro[cyclohexane-1,3′-indoline] derivatives with 6-(R)-hydroxy configuration. Whereas, the electron-donating N-protecting group furnished spiro[cyclohexane-1,3′-indoline] derivatives with 6-(S)-hydroxy configuration as the major product.

We initially investigated a catalytic domino reaction of 3-olefinic oxindole 1a and pentane-1,5-dial 2^12,13 using organocalatyst A (10 mol%) in THF at ambient temperature, under an aerobic atmosphere as shown in Scheme 1. The desired tandem Michael/aldol product 5a was isolated in 41% yield. As shown in Table 1, the ratio of diastereomers was moderate (3.5:1:0.5:0.3). However, the major product 5a showed excellent enantioselectivity (96% ee, Table 1, entry 1) in HPLC analysis.¹⁷ The overall process presumably proceeded through Michael reaction resulting in intermediate 3 followed by an intramolecular aldol reaction providing intermediate 4. Addition of an acidic co-catalyst, such as ortho-fluorobenzoic acid, trifluoroacetic acid or acetic acid, failed to improve the yield (Table 1, entries 2-4). We investigated other organic solvents such as CH₂Cl₂, toluene, N,N-dimethylformamide (DMF), hexane, and CH₃CN (entries 5-9). As it turned out that the choice of solvent was crucial to this Michael-Aldol tandem reaction and the best results were obtained using DMF as the solvent, affording 5a as the major diastereomer (6:1:0.5:0.4) in 80% yield and showed excellent enantioselectivity for the major diastereomer (>99% ee, entry 8). An attempt to perform the reaction at lower temperature resulted in low conversion (entry 10).

Scheme 1.

Michael/aldol cascade with oxindole 1a and dialdehyde 2.

Table 1.

Michael/Aldol Reaction Optimization^a

entry	solvent	T (h)	yieldb (%)	dr c	ee (%)d
1	THF	48	41	3.5:1:0.5:0.3	96
2e	THF	48	<10	nd	n. d.
3f	THF	48	<5	nd	n. d.
4g	THF	48	<5	n. d.	n. d.
5	CH2Cl2	48	<5	n. d.	n. d.
6	toluene	48	<5	n. d.	n. d.
7	hexane	48	<5	n. d.	n. d.
8	DMF	20	80	6:1:0.5:0.4	>99
9	CH3CN	48	<5	n. d.	n. d.
10h	DMF	48	16	7:1:0.5:04	99

^aUnless otherwise noted, the reaction was performed by employing 1a (0.1 mmol), dialdehyde 2 (50% in water, 0.3 mmol), and organocatalyst A (0.01 mmol) in the indicated solvent (0.5 mL) at room temperature. n. d. = not determined.

^bIsolated yield.

^cDetermined by ¹H NMR spectroscopy.

^dDetermined by HPLC of the major diastereomer.

^eOrtho-fluorobenzoic acid (0.01 mmol) was added.

^fTrifluoroacetic acid (0.01 mmol) was added.

^gAcetic acid (0.01 mmol) was added.

^hThis reaction was performed at 0 °C.

We subsequently explored substrate scope and limitations under the optimized reaction conditions (Figure 1). As evidenced by the results in Table 2, the reaction proceeded smoothly with both aromatic and aliphatic substituents to give desired products in high yields. The substituents on the aryl ring have little effect on the enantioselectivity of the major isomer. Monosubstituted phenyl methyleneindol-inones proceeded with excellent enantioselectivity ranging from 98 to 99% ee (entries 1-7). However, electronic properties of aryl substituents showed definite effect on diastereoselectivity. Electron-donating groups resulted in better diastereoselectivity. As shown, a high diastereomeric ratio of 7:1:0.5:0.4 was observed for 3-methoxyphenyl substituted methyleneindolinone, whereas a dr of 3:1:0.7:0.5 was observed for 3-nitro-phenyl substituted methyleneindolinone (entry 3 vs 7). Reactions with both heterocyclic and alkyl methyleneindolinones with aromatic substituents also proceeded with high yields and excellent enantioselectivity (entries 8 and 9). We further examined the cyclization reaction using an N-acetyl group in indole in place of an N-Boc protecting group. This resulted in product 5j in excellent yield with excellent enantioselectivity; however, the diastereomeric ratio (3:1:0.8:0.2; entry 10) was lower compared to the Boc-protected derivative (entry 1).

Figure 1.

Structure and enantioselectivity of major isomer

Table 2.

Enantioselective syntheses of Spirooxindoles ^a

Entry	R1	Pdt (Y%)b	Drc (ee %)d
1	Ph	5a (80)	6:1:0.5:0.4 (99)
2	4-Me-Ph	5b (82)	6:1:0.5:0.3 (99)
3	3-OMePh	5c (82)	7:1:0.5:0.4 (98)
4	2-OMe-Ph	5d (84)	5.5:1:0.5 (99)
5	3-Cl-Ph	5e (74)	5.2:1:0.5:0.5 (99)
6	4-NO2-Ph	5f (80)	3:1:0.5 (99)
7	3-NO2-Ph	5g (74)	3:1:0.7:0.5 (99)
8e	2-furyl	5h (80)	6:1:0.5:0.3 (98)
9e	Pr	5i (54)	6:1:0.7 (95)
10f	Ph	5j (90)	3:1:0.8:0.2 (95)

^aGeneral reaction conditions: oxindole (0.1 mmol), dialdehyde 2 (50% in water, 0.3 mmol), and organocatalyst A (0.01 mmol) in DMF (0.5 mL) at 23°C.

^bIsolated combined yield.

^cThe ee was determined by chiral-phase HPLC analysis of major diastereomer.

^dDetermined by ¹H NMR of crude product.

^eThe reaction was performed for 3 days.

^fOxindole is protected as NAc.

We then examined the effect of electron-donating N-protecting groups on the oxindole. Interestingly, this resulted in aldol ring closure with opposite hydroxyl stereochemistry as the major diastereomer. The results in Table 3 show various N-alkyl protecting groups, such as methyl, allyl and benzyl groups, provided products 5k-5m as the major diastereomers in excellent yields (70-84%) and excellent enantioselectivity for the major diastereomer (>99%, entries 1-3). Furthermore, both electron-donating (Me) and electron-withdrawing (Cl) substituents at the C4-position of the phenyl group afforded the corresponding products in high yields with excellent enantioselectivity for the major diastereomer (>99% ee, entries 4 and 5). Interestingly, the electron-withdrawing 4-NO₂ derivative furnished moderate diastereoselectivity (entry 6) compared to 4-Cl substituent (entry 4). However, the major isomer showed excellent enantioselectivity.

Table 3.

Syntheses of Spirooxindoles with N-alkyl groups^a

entry	R1	R2	Pdt(Y%) b	drc (ee %)d
1	Ph	Me	5k (70)	6.5:1:0.5 (99)
2	Ph	Bn	5l (82)	6:1:0.5:0.3 (99)
3	Ph	allyl	5m (84)	7:1:0.4 (99)
4	p-ClPh	allyl	5n (74)	6.5:1:0.3:0.2 (99)
5	p-MePh	allyl	5o (74)	7:1:0.5:0.5 (99)
6	pNO2Ph	allyl	5p (70)	2:1:0.5 (99)

^aGeneral reaction conditions: the reaction was performed by employing 3-olefinic oxindole (0.1 mmol), dialdehyde 2 (50% in water, 0.3 mmol), and organocatalyst A (0.01 mmol) in DMF (0.5 mL) at room temperature for 2 days.

^bIsolated yield.

^cDetermined by ¹H NMR of crude product.

^dThe ee was determined by chiral-phase HPLC analysis of major diastereomer.

We have also explored this Michael/aldol reaction with 3-methylpentane-1,5-dial 6¹⁸ as shown in Scheme 2. This reaction proceeded smoothly and afforded spirocyclic oxindole 5q with five consecutive stereogenic centers in 65% yield and excellent enantioselectivity (99% ee). Interestingly, the aldehyde functionality at the C3-position was isomerized from the axial to equatorial position leading to the formation of the more stable C2–C3 anti-product.

Scheme 2.

Reaction of oxindole 1a and 3-methylpentane-1,5-dial 6

We have carried out the reaction of oxindole 1a and dialdehyde 2 up to a gram scale providing 5a in 75% yield (Scheme 3). This product showed excellent enantioselectivity as well. The relative configurations of compounds were assigned by extensive NOE analyses of compounds 5a, 5d, 5o and 5q. The absolute configurations of the tandem Michael/aldol products were determined unambiguously through X-ray crystallographic analysis. As shown in Scheme 3, benzoate derivative 7 was prepared by reduction of 5a with NaBH₄ followed by reaction of the resulting alcohol with 4–nitrobenzoyl chloride in the presence of Et₃N in 56% yield over 2-steps. This was later recrystallized from methanol (23 °C). Subsequent single crystal X-ray crystallographic analysis supported the assignment of the relative and absolute stereochemistry shown in Figure 2 (see supporting information for the details of X-ray analysis).¹⁹

Scheme 3.

Synthesis of benzoate 7

Figure 2.

ORTEP drawing of 5l and 7

In conclusion, we have developed a diastereoselective organocatalytic Michael/aldol cascade reaction that provided convenient access to functionalized spirooxindoles with up to five consecutive stereogenic centers, including a spiro quaternary center. The products were obtained in excellent yield and the major diastereomer showed excellent enantioselectivity. In addition, depending upon our selection of N-protecting group on the oxindole, we were able to effectively control the stereochemical outcome of the hydroxyl center upon aldol ring closure. Further studies and applications of this methodology are the subject of current investigation in our laboratory.