A One-Pot, 3-Component, Domino Heck-aza-Michael Approach to Libraries of Functionalized 1,1-Dioxido-1,2-benzisothiazoline-3-acetic acids

Abstract

A sequential three-component synthesis of functionalized benzisothiazoline-3-acetic acid 1,1-dioxides utilizing a domino Heck-aza-Michael pathway is reported. This one-pot procedure rapidly assembles functionalized benzylsulfonamides, which undergo a palladium-catalyzed, domino, Heck-aza-Michael transformation in an experimentally straightforward manner. This attractive protocol has been utilized to synthesize three combinatorial sub-libraries (I-III) comprising a total of 95 compounds in high purities (≥95% for 75 compounds), yield and quantities.

Introduction

Advances in high-throughput screening and the need for new pharmaceutical leads have prompted the development of new protocols to generate diverse libraries of drug-like compounds. In recent years, a number of facilitated protocols that utilize both solid-phase and solution-phase chemistry have emerged to meet this challenge. Despite success in this area, there are limited examples of protocols that take advantage of cross-reaction functionality to allow domino/tandem processes to occur in a multi-component one-pot procedure.ⁱ

Recently, we reported a one-pot, sequential three-component approach towards the synthesis of 1,2-benzisothiazoline-3-acetic acid 1,1-dioxides (Scheme 1).ⁱⁱ The key step in this protocol was the utilization of a domino Heck-aza-Michael (HaM) reaction for both the mode of cyclization and as a pathway for the incorporation of an additional point of diversification.

Scheme 1.

Sequential three-component domino Heck-aza-Michael

Utilizing the (HaM) protocol, a small proof of concept demonstrative library of 1,2-benzisothiazoline-3-acetic acid 1,1-dioxides and subsequent derivatives was reported.ⁱⁱ Building on this work, the application of a domino, HaM protocol in the synthesis of three combinatorial sub-libraries (I-III) utilizing a variety of reaction platforms (Figure 1) is herein reported. Additionally a wider range of coupling partners was utilized to ultimately afford structural diversity around the central core. In order to maximize their potential drug-like properties, the subsequent libraries feature a multi-faceted design scheme as well as in-silico screening against Lipinski's rule of five criteria. The in-silico data ranges include a molecular weight range under 500 g/mol, no more than 5 hydrogen bond donors and/or 10 hydrogen bond acceptors, and a partition coefficient log P (clogP) less than 5.0.ⁱⁱⁱ Overall, the application of a domino Heck-aza-Michael (HaM) allows for rapid incorporation of functionality via the manipulation of the three individual components, allowing for the design of a library of diverse drug-like small molecules.

Figure 1.

Overview of the prepared libraries

Sultams (cyclic sulfonamide analogues) have emerged as important targets in drug discovery due to their extensive chemical and biological profiles.^iv Though not found in nature, a number of benzofused sultams have recently surfaced in the literature, which display potent activity across a variety of biological targets. Such reports include inhibition of a variety of enzymes, including COX-2 (Ampiroxicam),^v,vi HCV NS5b RNA-dependent RNA-polymerase,^vii HIV integrase,^viii cysteine proteases involved in the progression of maleria^ix and lipoxygenases.^x In addition, sultams have also shown antimycobacterial activity against M. Tuberculosis,^xi and inhibition of melanin-concentrating hormone [MCH] (Figure 2).^xii

Figure 2.

Representative biologically active sultams

This aforementioned biological profile is augmented by a number of inherent chemical properties possessed by both sultams and their sulfonamide precursors, including facile coupling/allylation pathways for sulfonamide and sultam formation, hydrolytic stability, polarity and their crystalline nature. Traditionally, sultams have been synthesized utilizing a variety of classical cyclization protocols such as Friedel-Craft,^xiii dianion,^xiv [3+2] cycloadditions,^xv, Diels-Alder reaction,^xvi and recently the application of an oxa/aza-Michael reaction.^xvii However, recently there have been a number of transition metal-catalyzed protocols reported utilizing ring-closing methathesis (RCM),^xviii Heck,^xix as well as Au-,^xx Cu-^xxi and Rh-catalyzed^xxii cyclization protocols for the generation of diverse sultams.

Results and Discussion

A 56-member library I was initially designed to expand on the prototype library previously reported,ⁱⁱ demonstrating the capability of the HaM protocol in a library format. The method allows for the generation of sultams with three points of diversification starting from commercially available α-bromobenzenesulfonyl chlorides coupled with a range of aromatic, cyclic and alkyl amines (Figure 3).^xxiii,xix

Figure 3.

Representative coupling partners for Heck-aza-Michael

Specific combinations of the three-components were chosen to evaluate the robustness of this protocol to peripheral functionality. In addition, Lipinski's rule of five also guided in-silico efforts in the selection of combinations. Based on previous work, it was anticipated that these substrates would be well tolerated under the reaction conditions, carrying out the preparation of library I in 1-dram vials on an aluminum reaction block.^xxiv To this effect, α-bromobenzylsulfonyl chlorides (S1-3) were coupled with amines (A1-9) and stirred for 2 hours at room temperature. After such time, Et₃N, Bu₄NCl, Pd₂(dba)₃·CHCl₃ and the corresponding Michael acceptors (M1-4) were added to the reaction mixture, which was heated to 110 °C and subjected to workup after 14 hours. Workup consisted of removal of DMF, suspension of the crude reaction mixture in EtOAc and filtration through a SiO₂ SPE to remove inorganic salts and spent palladium. The crude material was analyzed by HPLC (UV 214 nm) and submitted to purification by mass-directed fractionation (MDF) to yield the anticipated products in modest to good yield and high purity (Table 1).^xxv

Table 1.

Summary of Library I results

entry	R1	R2NH2	R3	yield (%)	purity (%)	product
1	S1	A1	M1	44	100	1
2	S1	A6	M1	48	97	2
3	S1	A7	M1	53	98	3
4	S1	A5	M2	22	98	4
5	S1	A6	M2	22	95	5
6	S1	A7	M2	74	100	6
7	S1	A3	M3	57	100	7
8	S1	A4	M3	64	100	8
9	S1	A5	M3	35	100	9
10	S1	A6	M3	34	95	10
11	S1	A7	M3	60	98	11
12	S1	A3	M4	42	100	12
13	S1	A4	M4	29	92	13
14	S1	A5	M4	40	99	14
15	S1	A6	M4	57	90	15
16	S2	A1	M1	12	99	16
17	S2	A2	M1	30	99	17
18	S2	A5	M1	40	90	18
19	S2	A7	M1	25	96	19
20	S2	A8	M1	75	93	20
21	S2	A9	M1	69	90	21
22	S2	A1	M2	28	100	22
23	S2	A2	M2	46	96	23
24	S2	A3	M2	87	94	24
25	S2	A4	M2	57	95	25
26	S2	A5	M2	56	98	26
27	S2	A6	M2	40	94	27
28	S2	A7	M2	69	97	28
29	S2	A8	M2	72	97	29
30	S2	A9	M2	54	96	30
31	S2	A3	M3	58	100	31
32	S2	A4	M3	64	97	32
33	S2	A5	M3	13	93	33
34	S2	A7	M3	54	100	34
35	S2	A8	M3	50	99	35
36	S2	A9	M3	45	100	36
37	S3	A1	M1	25	96	37
38	S3	A2	M1	21	97	38
39	S3	A3	M1	53	99	39
40	S3	A4	M1	51	100	40
41	S3	A7	M1	52	99	41
42	S3	A8	M1	31	91	42
43	S3	A9	M1	42	96	43
44	S3	A1	M2	50	95	44
45 d	S3	A3	M2	5	100	45
46 d	S3	A4	M2	7	100	46
47 d	S3	A6	M2	5	87	47
48 d	S3	A7	M2	6	97	48
49 d	S3	A8	M2	9	96	49
50 d	S3	A9	M2	9	93	50
51	S3	A2	M3	27	97	51
52	S3	A3	M3	42	93	52
53	S3	A4	M3	46	95	53
54	S3	A7	M3	35	92	54
55	S3	A8	M3	44	92	55
56	S3	A9	M3	40	99	56

^aReaction conditions: 1 (0.273 mmol), Pd₂(dba)₃.CHCl₃ (2 mol %), methyl acrylate (0.82 mmol), Bu₄NCl (0.273 mmol) in DMF at 110 °C for 14 h.

^bPurified by an automated preparative reverse phase HPLC (detected by mass spectroscopy).

^cPurity was determined by HPLC with peak area (UV) at 214 nm.

^dWhen repeated in single reaction formation, good yields (50-80%) were achieved as predicted.

Overall, a total of 56 reactions afforded product in variable yield and purity, validating the scope and economy of the HaM strategy. Specifically, out of the 56 reactions carried out, 44 had a final purity of 95% or greater with reactions yielding good overall mass recovery.^xxvi Having established the viability of the HaM protocol in the generation of libraries, we set out to design a library (Library II) of 1,2-benzisothiazoline-3-acetic acid 1,1-dioxides utilizing the Bohdan MiniBlock® platform. Under this premise, a library of compounds was prepared in parallel using a 26-member Bohdan MiniBlock. In an attempt to design more drug-like molecules a new collection of amines (A10-15) was employed (Table 2). In addition, reactions were carried out using stock solutions to streamline the process thereby granting it the potential for future automation. As with sub-library I, crude compounds were analyzed by HPLC (UV 214 nm) and submitted to purification by MDF. Overall, with the exception of compounds 60 and 81, all reactions worked with good yield and high purity. Notably, 20 out of the 24 reactions had a final purity of 95% or greater. Having validated the methodology, the protocol was implemented for the generation of 1,2-benzisothiazoline-3-acetic acid 1,1-dioxides 84-95 bearing both saturated and unsaturated side chains as previously reported (Table 3). The addition of one more point of diversification was accomplished by utilizing commercially available 2,5-dibromosulfonyl chloride. In this regard, 12 parallel reactions were carried out in a Radley Carousel®, utilizing stock solutions for the quick generation of the desired products in a sequential three-component coupling. Crude material was analyzed via HPLC (UV 214 nm) and submitted to purification by MDF. As expected, all reactions resulted in good yields and high purities yielding the desired compounds with good mass recovery.

Table 2.

Representative components and products synthesized using Bohdan MiniBlock platform

entry	R1	R2NH2	R3	yield (%)	purity (%)	product
1	S1	A10	M1	92	98	57
2	S1	A11	M4	12	98	58
3	S1	A12	M3	44	100	59
4	S1	A13	M4	6	100	60
5	S1	A14	M2	56	100	61
6	S1	A15	M2	68	100	62
7	S2	A10	M1	93	99	63
8	S2	A10	M4	82	99	64
9	S2	A11	M3	16	100	65
10	S2	A12	M1	84	90	66
11	S2	A12	M3	16	97	67
12	S2	A15	M2	82	100	68
13	S3	A10	M1	40	100	69
14	S3	A10	M3	32	95	70
15	S3	A10	M4	52	100	71
16	S3	A11	M3	18	91	72
17	S3	A12	M1	36	92	73
18	S3	A12	M4	20	100	74
19	S3	A13	M1	30	100	75
20	S3	A13	M3	34	93	76
21	S3	A13	M4	16	100	77
22	S4	A10	M1	32	100	78
23	S4	A11	M1	30	99	79
24	S4	A11	M3	24	97	80
22	S4	A12	M3	8	100	81
23	S4	A13	M1	22	100	82
24	S4	A13	M4	16	100	83

^a Reaction conditions: 1 (0.136 mmol), Pd₂(dba)₃.CHCl₃ (2 mol %), Michael acceptor (0.40 mmol), Bu₄NCl (0.136 mmol) in DMF at 110 °C for 14 h. ^b Purified by an automated preparative reverse phase HPLC (detected by mass spectroscopy). ^c Purity was determined by HPLC with peak area (UV) at 214 nm.

Table 3.

Representative components and products synthesized using the Radley Carousel® platform

entry	R2NH2	R3	yield (%)	purity (%)	compound
1	A7	M1	18	100	84
2	A7	M2	45	93	85
3	A7	M3	62	90	86
4	A7	M4	86	100	87
5	A9	M1	69	100	88
6	A9	M2	90	100	89
7	A9	M3	86	99	90
8	A9	M4	59	100	91
9	A16	M1	78	100	92
10	A16	M2	81	100	93
11	A16	M3	60	100	94
12	A16	M4	90	90	95

^a Reaction conditions: 1 (0.136 mmol), Pd₂(dba)₃.CHCl₃ (2 mol %), methyl acrylate (0.40 mmol), Bu₄NCl (0.136 mmol) in DMF at 110 °C for 14 h. ^b Purified by an automated preparative reverse phase HPLC (detected by mass spectroscopy). ^c Purity was determined by HPLC with peak area (UV) at 214 nm.

In conclusion, the successful demonstration of a sequential three-component synthesis of functionalized benzisothiazoline-3-acetic acid 1,1-dioxides utilizing a domino Heck-aza-Michael pathway has been accomplished. This one-pot, three step procedure, rapidly assembles functionalized benzylsulfonamides, which undergo a palladium-catalyzed domino Heck-aza-Michael transformation in an experimentally straightforward manner. This protocol was demonstrated on a variety of platforms (Reaction blocks, Bohdan MiniBlock® and Radley Carousel®) producing overall three combinatorial sub-libraries (I-III) comprising a total of 95 pure compounds in high purities (≥95% for 75 compounds) and good quantities. The evaluation of biological activities of the compounds reported herein in high-throughput screens is currently underway. Future efforts will continue to focus on the development of new methodology for the synthesis of diverse, functionalized libraries and their biological evaluation.

Experimental Section

General procedures: All air and moisture sensitive reactions were carried out in flame- or oven-dried glassware under argon atmosphere using standard gas tight syringes, cannula, and septa. Stirring was achieved with oven-dried, magnetic stir bars. CH₃CN was purified by passage through the Solv-Tek purification system employing activated Al₂O₃ (Grubbs, R. H.; Rosen, R. K.; Timmers, F. J. Organometallics 1996, 15, 1518-1520). Et₃N was purified by passage over basic alumina and stored over KOH. Flash column chromatography was performed with SiO₂ from Sorbent Technology (30930M-25, Silica Gel 60A, 40-63 um). Thin layer chromatography was performed on silica gel 60F254 plates (EM-5717, Merck). Deuterated solvents were purchased from Cambridge Isotope laboratories. ¹H and ¹³C NMR spectra were recorded on a Bruker DRX-400 NMR spectrometer operating at 400 MHz and 100 MHz respectively; or a Bruker Avance operating at 500 MHz and 125 MHz respectively. High-resolution mass spectrometry (HRMS) and FAB spectra were obtained in one of two manners: (i) on a VG Instrument ZAB double-focusing mass spectrometer and (ii) on a LCT Premier Spectrometer (Micromass UK Limited) operating on ESI (MeOH). All library syntheses using block technology were performed using a 24-position Mettler-Toledo Bohdan MiniBlock XT under an argon atmosphere in oven-dried Autochem 17 × 100 mm round bottom tubes. Parallel evaporations were performed using a GeneVac EZ-2 plus evaporator. Automated preparative reverse-phase HPLC purification was performed using a Waters 2767 Mass-Directed Fractionation system (2767 sample manager, 2525 Binary Pump, 515 Make-up pump) with a Waters ZQ quadrapole spectrometer and detected by UV (270 nm, Waters Xterra MS C-18 column, 19×150 mm, elution with the appropriate gradient of CH₃CN in pH 9.8 buffered aqueous ammonium formate at 18 mL min^-1 flow rate). Purity was determined by reverse-phase HPLC with peak area (UV) at 214 nm using a Waters Alliance 2795 system (Waters Xterra MS C-18 column, 4.6×150 mm, elution with a linear gradient of 5% CH₃CN in pH 9.8 buffered aqueous ammonium formate to 100% CH₃CN at 1.0 mL/min flow rate).

General procedure (A) for the synthesis of Library I (1-56) on Reaction Blocks in 1 dram vials

Into a 1-dram vial was added amine (0.237 mmol), Et₃N (0.546 mmol) and dry DMF (0.60 mL) and the reaction was stirred at RT for 15 minutes. After such time, α-bromobenzenesulfonyl chlorides (0.237 mmol) were added and the reaction was stirred for 2 hrs. To the reaction vial was added Et₃N (0.546 mmol), Bu₄NCl (0.237 mmol), Pd₂(dba)₃•CHCl₃ (2 mol%) and dry DMF (1.4 mL). After stirring for 5 min at RT, Michael acceptor (0.819 mmol) was added and the reaction vial was placed immediately into a preheated reaction block. The reaction was stirred at 110 °C for 14 hrs after which time the reaction was cooled and concentrated under reduced pressure. The crude was suspended in EtOAc, filtered through a SiO₂ SPE and analyzed by HPLC (UV 214 nm). Crude material with purity below 90% was submitted to purification by MDF.

General procedure (B) for the synthesis of Library II (57-83) in a Bohdan MiniBlock

Into a MiniBlock reaction tube, was added a stock solution of amine (0.136 mmol) in dry DMF (0.10 mL) followed by Et₃N (0.273 mmol) in dry DMF (0.10 mL) and reaction was stirred at RT for 15 minutes. A stock solution of α-bromobenzenesulfonyl chloride (0.136 mmol) in dry DMF (0.10 mL) was added and the reaction was stirred for 2 hrs. After such time, a stock solution of Et₃N (0.273 mmol), Bu₄NCl (0.136 mmol) and Pd₂(dba)₃•CHCl₃ (2 mol%) in dry DMF (0.7 mL) was added to the reaction mixture. The MiniBlock was then heated to 110 °C and the Michael acceptor (0.410 mmol) was added. After stirring at 110 °C for 14 hrs, the crude reaction was cooled to RT and concentrated under reduced pressure. The crude was suspended in EtOAc, filtered through a SiO₂ SPE and analyzed by HPLC (UV 214 nm). Crude material with purity below 90% was submitted to purification by MDF.

General procedure (C) for the synthesis of Library III (84-95) in a Radleys Carousel

Into a reaction tube contained within a 12-port Radley Carousel®, was added a stock solution of amine (0.136 mmol) in dry DMF (0.10 mL) followed by Et₃N (0.273 mmol) in dry DMF (0.10 mL) and the reaction was stirred at RT for 15 minutes. A stock solution of α-bromobenzenesulfonyl chloride (0.136 mmol) in dry DMF (0.10 mL) was added and the reaction was stirred for 2 hrs. After such time, a stock solution of Et₃N (0.273 mmol), Bu₄NCl (0.136 mmol) and Pd₂(dba)₃•CHCl₃ (2 mol%) in dry DMF (0.7 mL) was added to the reaction mixture. The MiniBlock was then heated to 110 °C after which time, the Michael acceptor (0.410 mmol) was added. After stirring at 110 °C for 14 hrs, the crude reaction was cooled to RT and concentrated under reduced pressure. The crude was suspended in EtOAc, filtered through a SiO₂ SPE and analyzed by HPLC (UV 214 nm). Crude material with purity below 90% was submitted to purification by mass-directed fractionation (MDF).