Catalytic Stereoselective Synthesis of Diverse Oxindoles and Spirooxindoles from Isatins

Abstract

A strategy for the efficient two-step synthesis of triazole derivatives of oxindoles and spirooxindoles is presented. Using a common set of N-propargylated isatins, a series of mechanistically-distinct stereoselective reactions with different combinations of nucleophiles and catalysts provide access to diverse hydroxy-oxindoles, spiroindolones, and spirocyclic oxazoline structures. The resulting N-propargylated oxindoles are then converted to triazoles using copper-catalyzed azide-alkyne cycloaddition (CuAAC) reactions. Overall, this strategy affords a 64-member pilot-scale library of diverse oxindoles and spirooxindoles.

Introduction

Complex functionalized molecules are important compounds of interest for biological probes and as new molecules for pharmaceutical lead discovery. Oxindole and spirooxindole scaffolds have generated considerable synthetic interest due to their occurrence in diverse natural products and notable biological activity.¹ In a recent discovery, spiroindolone NITD609 demonstrated nanomolar activity as a therapeutic agent that kills the blood stage of Plasmodium falciparum and has single-dose efficacy in a rodent malaria model (Figure 1).² Various hydroxy-oxindoles scaffolds also demonstrate important biological activity, such as Convolutamydine A, a natural product with potent activity against leukemia cells.³ Substituted isatin (indole-2,3-diones) scaffolds have also shown promising examples of biological activity.⁴ For example, isatin 1 is a potent inhibitor of SARS CoV 3C-like proteases.⁵ Initially driven by efficient synthetic methods, the 1,2,3-triazole has now emerged as a heterocycle of biological interest in drug discovery and medicinal chemistry programs.⁷ For example, triazole 2 shows activity against tuberculosis strain H37RV.⁶ The significant biological activities observed for oxindoles and triazoles emphasizes the need to develop efficient synthetic strategies to access these scaffolds and increase structural diversity for drug discovery and medicinal chemistry programs.

Figure 1.

Biologically active oxindoles and triazoles

Previous work from our laboratory has demonstrated several methods of catalytic activation of the isatin dicarbonyl for efficient and selective nucleophilic additions and spirocyclizations at the 3-position.⁸ The strategy we envisioned utilizes a common set of N-propargylated isatins 3 to access diverse oxindole scaffolds through a series of mechanistically-distinct nucleophilic addition pathways. Each resulting oxindole scaffold contains an alkyne group that provides further opportunities for structural diversification with triazole heterocycles using the copper-catalyzed azide-alkyne cycloaddition (CuAAC) reaction. We also envisioned that the N-propargylated isatins 3 would be directly utilized in the CuAAC reaction as a fifth scaffold because triazole-containing isatins have not previously been reported or evaluated for biological activity. Here we describe the realization of this strategy for rapid access (two synthetic steps) to a pilot-scale library of diverse triazole-containing hydroxy-oxindoles 4–6, spiroindolones 7, isatin-triazoles 8, and spirooxazolines 9–10 emanating from a common set of propargylated isatins 3 (Scheme 1). Overall, our two-step strategy affords a library of 64 oxindole and spirooxindole compounds, including 15 core scaffolds and 49 triazole derivatives. The efficient stereoselective syntheses of complex heterocycles combining both oxindole and triazole motifs have not been described previously. Based on the breadth of biological activity known for isatins, oxindoles and spirooxindoles, these densely functionalized heterocycles should serve as important biological probes for chemical biology.

Scheme 1.

Outline of Synthetic Strategy towards Substituted Triazole-containing Oxindoles

Results and Discussion

A key feature in this strategy is the regio-, diastereo- and enantioselective synthesis of the oxindole and spirooxindole scaffolds (Scheme 2). Based on the nucleophile component (11) utilized (Scheme 1), four oxindole scaffolds were selected for this library: hydroxy-oxindoles 13–15 are prepared enantioselectively using a chiral Lewis acid catalyst (19);^8b spiroindolones 16 are prepared enantioselectively using a chiral Brönsted acid catalyst (20);^8c and the 2-oxazoline and 3-oxazoline spirocycles 17 and 18 are each prepared diastereo- and regioselectively using a titanium(IV) Lewis acid catalyst (Scheme 2).^8a For some scaffolds, isatins (3) are selected based on substitution patterns that ensure high selectivity.

Scheme 2.

Nucleophilic Addition Reactions to Access Oxindoles and Spirooxindoles.

First, a series of 3-substituted-3-hydroxy-oxindole scaffolds 13–15 were accessed using scandium(III)-catalyzed enantioselective additions with representative π-nucleophiles: N-methylindole (11a), 2-methallylsilane (11b), and N,N-dimethyl-m-anisidine (11c).^8b Scandium(III) complexes formed with the 2,6-bis[(3aS,8aR)-3a,8a-dihydro-8H-indeno[1,2-d]oxazolin-2yl]pyridine ligand (e.g. 19) are known to be effective chiral Lewis acid catalysts with good chelating potential.⁹ As outlined in Table 1, isatins 3b–c were utilized to afford hydroxy oxindoles 13–15 with high yields (78–97%) and enantioselectivity (85–99% ee). All reactions were performed at room temperature, with the exception of entries 1 and 4, which were performed at −20 °C due to the high reactivity of the nucleophile. In the case of the methallylsilane (11b, entries 2 and 5), the reaction is run in acetonitrile with TMSCl and NaSbF₆ as additives to increase the efficiency of the reaction and promote in situ deprotection of any resulting OTMS product so that the hydroxy-oxindole products 13–15 are obtained exclusively.^10,11

Table 1.

Enantioselective Synthesis of Substituted Hydroxy-Oxindole Scaffolds

entry	R	isatin	nucleophile	product	yielda	%eeb
1c	5-F	3b	11a	13b	97	98
2d	5-F	3b	11b	14b	79	87
3	5-F	3b	11c	15b	97	96
4c	4-Cl	3c	11a	13c	78	86
5d	4-Cl	3c	11b	14c	90	94
6	4-Cl	3c	11c	15c	97	99

^aIsolated yields.

^bDetermined using HPLC analysis with chiral stationary phase.

^cReaction performed at −20 °C.

^dReaction was performed using 3.0 equiv of TMSCl and 0.1 equiv of NaSbF₆ in MeCN.

Using an asymmetric Pictet-Spengler-type spirocyclization reaction,¹² three spiroindolone scaffolds were prepared upon acid-catalyzed spirocyclization of 5-methoxytryptamine (11d) with isatins (Table 2).^8c,13 We have previously shown that this reaction is efficiently catalyzed using Lewis acidic metal salts, thioureas, or BINOL-derived chiral phosphoric acids. The 5-fluorospiroindolone 16b was prepared using a BINOL-derived phosphoric acid catalyst to demonstrate the enantioselective synthesis and evaluate the retention of enantiomeric excess in the CuAAC reaction. Using (S)-3,3′-bis(2,4,6-triisopropylphenyl)-1,1′-binaphthyl-2,2′-diyl hydrogenphosphate (20) as the catalyst, fluorospiroindolone 16b was attained in excellent yield (86%) and high enantioselectivity (84% ee) (Table 2, entry 2).¹¹ Using either Sc(OTf)₃ or 1,3-bis(3,5-bis(trifluoromethyl)phenyl)thiourea provides an efficient (and less expensive) catalyst for the preparation of racemic spiroindolones (Table 2, entries 1 and 3). Isatins 3b and 3f were selected due to the therapeutic relevance of fluorinated molecules as well as the efficient performance of these isatins in the Pictet-Spengler reaction.^{2b, 14}

Table 2.

Enantioselective Synthesis of Spiroindolone Scaffolds

entry	R	isatin	product	catalyst	yield (%)a	%eeb
1	H	3a	16a	Sc(OTf)3	89	--
2	5-F	3b	16b	20	86	84
3	7-F	3f	16f	Thiourea	92c	--

^aIsolated yield.

^bDetermined using HPLC analysis with chiral stationary phase.

^cReaction was performed in DCM solvent.

A series of spirooxindole oxazoline scaffolds were prepared using the titanium(IV)-catalyzed regio- and diastereoselective addition and spirocyclization of 5-methoxyoxazoles to isatins.^{8a, 15} Substitution at the 4-position of the oxazole controls the regiochemistry, with the 4-isopropyloxazole (11f) giving rise to the 3-oxazoline scaffold 18 and the 4-H oxazole (11e) leading to the 2-oxazoline scaffold 17.^{8a, 11, 16} In general, the oxazole addition to propargyl isatins 3a–d proceeded with high regioand diastereoselectivity (>99% rr, up to 99% dr) and in high yields (up to 95%); however, in two cases (17b and 18c) the isolated yield was low due to the presence of by-products that proved to be difficult to separate by column chromatography. All reactions proceeds with high diastereoselectivity, but the 4-chloroisatin 3c is particularly effective for dictating high diastereoselectivity in the formation of 2-oxazolines, such as 17c.

Triazole Synthesis

This collection of diverse oxindole and spirooxindole scaffolds contains a common feature through the N-propargyl group of each scaffold, which can be further diversified using azide-alkyne cycloaddition chemistry. We set out to compare several variations of CuAAC reactions for this collection of N-propargylated oxindoles. We began by exploring one-pot reaction conditions that utilize in situ generated azides, prepared from the corresponding amine using the shelf-stable diazotransfer reagent 1H-imidazole-1-sulfonyl azide (21) (Scheme 3, eq 1).¹⁷ Unfortunately, this method did not work for these oxindole substrates due to side reactions and decomposition of starting material under the diazo transfer conditions. Several solvent combinations (THF, MeOH/THF, DCM), copper reagents, and conditions (i.e. Cu(I) vs. Cu(II) reduction by sodium ascorbate in situ) were explored, but only minimal triazole formation was observed for the one-pot procedure. Next, we investigated a route utilizing in situ generated aryl azides prepared from commercially-available aryl iodides with sodium azide in the presence of copper iodide (Scheme 3, eq 2).¹⁸ This method was also considered to be attractive because it provides access to diverse aryl triazoles, including amino-substituted triazoles that are not available when preparing azides from amines. This method has previously only been reported for less-functionalized substrates, but here the conditions generally afforded low yields (33–40%) for the more complex oxindole substrates. Low yields are attributed to the formation of undesired polymericside products.

Scheme 3.

Methods Compared for the Synthesis of Triazoles with Propargylated Isatins

Ultimately, we utilized a method applicable for the synthesis of both alkyl- and aryl-substituted triazoles using purified azides that were prepared in one-step from the corresponding amine diazotransfer reagent 21 (Table 4).¹⁹ Azides 12a–f were selected to include indolyl, pyridyl, ester, and phenyl derivatives in order to provide a variation in molecular weights, hydrogen bonding donor and acceptor properties, and capabilities for π-interactions. Azides 12a–e were each prepared from the amine using diazotransfer reagent 21, and azide 12f was prepared by direct displacement of a chloride with sodium azide.²⁰ After purification and isolation, azides 12a–f were then allowed to react directly with the N-propargyl-oxindoles or spirooxindoles in the presence of catalytic Cu(I) iodide (10 mol %) and di-isopropylethylamine (15 mol %) in THF to afford triazoles 4–10.²¹ A representative combination of N-propargyl oxindole scaffolds and azide components were evaluated to ascertain the scope of compounds accessible using this strategy (Figure 2). Overall, these reactions proceeded with high yields (> 70%) for 36 of 45 triazoles prepared, and only 8 of the 45 reactions afforded less than 66% yield (Table 4). These low yields are attributed to lower conversion and unreacted starting material for the given reaction time. Of practical consideration, the CuAAC reactions of isatins (entries 20–27) proceeded with excellent yields and were particularly clean and easy to purify.

Table 4.

Synthesis of Triazole-functionalized Isatins, Oxindoles, and Spirooxindoles

entry	R	isatin/oxindole	azide (R1)	producta	yield (%)b
1	5-F	13b	12a	4ba	78
2	5-F	13b	12b	4bb	80
3	5-F	13b	12c	4bc	99
4	5-F	14b	12a	5ba	64
5	5-F	14b	12b	5bb	71
6	5-F	15b	12a	6ba	56
7	5-F	15b	12b	6bb	99
8	4-Cl	13c	12a	4ca	40
9	4-Cl	13c	12b	4cb	70
10	4-Cl	14c	12a	5ca	94
11	4-Cl	14c	12b	5cb	58
12	4-Cl	15c	12a	6ca	91
13	4-Cl	15c	12b	6cb	94
14	4-Cl	15c	12c	6cc	58
15	H	16a	12c	7ac	95
16	H	16a	12f	7af	95
17	5-F	16b	12b	7bb	62
18	5-F	16b	12c	7bc	97
19	7-F	16f	12a	7fa	70
20	H	3a	12d	8ad	69
21	H	3a	12a	8aa	99
22	H	3a	12e	8ae	95
23	4-Cl	3c	12a	8ca	55
24	4-Cl	3c	12f	8cf	99
25	4-Cl	3c	12c	8cc	97
26	5-Br	3d	12a	8da	74
27	5-OMe	3e	12a	8ea	99
28	5-F	17b	12b	9bb	74
29	4-Cl	17c	12a	9ca	88
30	4-Cl	17c	12b	9cb	94
31	4-Cl	17c	12c	9cc	88
32	4-Cl	17c	12f	9cf	87
33	4-Cl	17c	12d	9cd	99
34	H	18a	12a	10aa	92
35	H	18a	12b	10ab	76
36	H	18a	12f	10af	70
37	H	18a	12d	10ad	90
38	5-F	18b	12a	10ba	80
39	5-F	18b	12b	10bb	74
40	5-F	18b	12c	10bc	70
41	5-F	18b	12e	10be	94
42	5-F	18b	12d	10bd	74
43	4-Cl	18c	12b	10cb	66
44	4-Cl	18c	12d	10cd	93
45	5-Br	18d	12a	10da	99

^aAll reactions performed with CuI (10 mol %) and DIPEA (15 mol %) in THF from 12–24 h.

^bIsolated yield.

Figure 2.

A) Representative final oxindole and spirooxindole products.

Because hydroxy-oxindoles 13–15 were prepared as enantiomerically-enriched compounds using asymmetric catalysis, it is especially important to demonstrate that the enantomeric excess of these compounds was maintained during the CuAAC reaction since these compounds are known to form stabilized carbocations under metal- and acid-catalyzed conditions.²² The enantiomerically-enriched N-propargylated 3-hydroxy-oxindoles were subjected to the CuAAC conditions with azides 12a–c, e (Table 4, entries 1–14). The enantiomeric excess of the triazole-containing 3-hydroxy3-indolyl-oxindole 4ba was confirmed by HPLC analysis using chiral stationary phase by comparison to the triazoles produced by racemic synthesis.²³ Due to the similarity of hydroxy-oxindoles 13–15, the retention of enantiomeric excess for 4ba was used as a model for retention of enantioselectivity for all triazole products. Enantiomerically-enriched spiroindolone 16b was used to demonstrate that the enantiomeric excess for this class of spirooxindoles is retained under CuAAC reaction conditions (entry 17).^23b

The spirocyclic oxazolines 17 and 18 afforded spirocyclic triazoles 9 and 10 in excellent yields (Table 4, entries 28–45); however, the reversed order of the reaction sequence was also investigated. For spirooxazoline-triazole 9, the titanium(IV)-catalyzed addition and spirocyclization of 5-methoxyoxazole 11e can also be performed using a triazole-containing isatin (derived from azide 12a) in 71% yield and maintaining high diastereoselectivity (97:3 dr). However, the success of the spirocyclization is dependent on the nature of the triazole substrate (see Supporting Information). Although the reaction proceeded successfully with the isatin triazole derived from p-methoxy phenyl azide 12a, the use of an isatin triazole derived from azide 12b did not undergo spirocyclization, presumably due to interactions between the indole ring and the titanium catalyst.

Although our earlier investigations had shown that the one-pot reaction using aryl iodides proceeded with low yield, we briefly explored this approach to incorporate additional triazole diversity (Table 5). For example, this one-pot method allows the synthesis of triazole isatin 8eh containing an amino group, which cannot be accessed with the previous route described above (Table 4). This CuAAC strategy uses 10 mol % of CuI and 15 mol % of DMEDA with one equivalent of aryl-iodide and one equivalent of the alkyne in DMF.¹⁸ Triazoles 8cg and 8eh were obtained when isatins 3c,e were subjected to the reaction with aryl-iodides; however, low yields were observed (Table 5, entries 1–2). Similarly, with hydroxy-oxindoles 14 and 15, the reaction was sluggish with catalytic amounts of copper, also affording low yields of the triazole products. When the amount of copper was increased to stoichiometric amounts, several unidentified side-products were formed and the increase in yield was negligible. In order to reduce the amount of side products and increase yields, conditions were modified to use a stoichiometric amount of copper(II) reagent with sodium ascorbate to generate the catalytic Cu(I) species in situ. Even with these optimized conditions, the yield of pyridinyl triazoles remained low (Table 5, entries 3–5). This result is in contrast to the reaction of pyridyl azide 12e, which was successfully utilized and afforded triazole products with excellent yields (Table 4, entries 22 and 41). However, this method can provide additional interesting compounds for biological screening.

Table 5.

Synthesis of Triazole-functionalized Oxindoles and Isatins from Aryl-iodides

entry	R1	isatin/oxindole	Ar-N3a	product	yield (%)b
1c	4-Cl	3c	12g	8ca	23
2c	5-OMe	3e	12h	8eh	35
3d	5-F	14b	12i	5bi	28
4d	4-Cl	14c	12i	5ci	40
5d	4-Cl	15c	12i	6ci	19

^aAzide generated in situ from the corresponding aryl-iodide.

^bIsolated yield.

^cReactions performed using CuI (10 mol %), and DMEDA (15 mol %) according to Scheme 3, eq 3.

^dReaction performed with 1 equiv of CuSO₄•5H₂O, and 1.1 equiv. of sodium ascorbate.

An analysis of the molecular properties and shape for this collection of oxindole compounds indicates desirable properties and diversity for high-throughput screening and the discovery of pharmaceutical leads or biological probes. Molecular properties for all compounds were calculated (see Supporting Information) and a summary of the average values are provided in Table 6. The majority of compounds and average values are within accepted ranges for the development of lead compounds.²⁴ The nucleophile and azide building blocks selected here afford a collection of compounds with molecular weights ranging from 259–721, and calculated partition coefficients (cLogP) values with a range of 0.56 to 5.38, based on the XLogP method of Wang.²⁵ The molecular weights span a range that includes molecular weights appropriate for lead-like molecules, as well as access to higher molecular weight compounds, which could prove useful as biological probes for the disruption of protein-protein interactions.²⁶ In order to visualize the molecular shape diversity, we generated a scatter plot based on the principal moments of inertia (PMI) ratios (Figure 3), a method developed by Sauer and Schwarz.²⁷ This method classifies the molecular shape into three categories: rod (acetylene), disk (benzene), or spherical (adamantine). Since several conformations of a compound are capable of binding to a biological target, a collection of 3D-conformations ≤3 kcal/mol from the minimum energy conformer are represented. This shape analysis also includes the structures of the known biologically active compounds in Figure 1 for comparison.

Table 6.

Average values of Molecular Properties

property	oxindole cores (n = 15)	isatin triazoles (n = 9)	oxindole triazoles (n = 40)
MW	399	353	551
XLogPa	3	2	3
TPSAb	67	81	109
rotatable bonds	4	4	8
HBA	3	4	5
HBD	1	0	1
# of rings	2	3	4

^aXLogP is used for the calculated partition coefficients (cLogP) values based on the method of Wang, see ref 25.

^bTopological Polar Surface Area (TPSA) based on the method of Ertl, see ref 28.

Figure 3.

Scatter plot with principal moments of inertia (PMI) ratios plotted to compare the molecular shape diversity of oxindole core structures (blue), triazole-containing isatins and oxindoles (red), and known biologically-active compounds from Figure 1 (green). For each compound, PMI ratios were calculated for all minimum energy conformers ≤3 kcal/mol from the global minimum.

In conclusion, we have developed an efficient enantio- and diastereoselective synthetic strategy to access a diverse 64-compound pilot-scale library including 15 oxindole scaffolds and 49 triazole containing-oxindoles and isatins. We demonstrate that enantiomeric excess resulting from the catalytic asymmetric synthesis of oxindoles and spirooxindoles is retained upon further functionalization with the CuAAC reaction, thus providing efficient methods to prepare libraries of enantiomerically-enriched spirocyclic compounds. The nucleophile and azide building blocks selected here afford a collection of compounds with diversity that is appropriate for high-throughput screening and the discovery of pharmaceutical leads or biological probes. All of the compounds in this report have been submitted to the NIH Molecular Libraries Small Molecule Repository for biological screening.

Table 3.

Regio- and Stereoselective Spirocyclization to Afford 2- and 3-spirooxazoline Scaffolds

entry	R	isatin	oxazole	product	yield (%)a	drb
1	5-F	3b	11e	17b	55c	90:10
2	4-Cl	3c	11e	17c	74	99:1
3	H	3a	11f	18a	77c	90:10
4	5-F	3b	11f	18b	87	93:7
5	4-Cl	3c	11f	18c	51c	95:5
6	5-Br	3d	11f	18d	95	99:1

^aIsolated yield of major diastereomer.

^bDetermined by analysis of ¹H NMR spectroscopy of crude reaction mixture.

^cYields sacrificed for purity due to the presence of by-products which proved to be difficult to separate by column chromatography (conversion ≥80% by TLC).