Isoxazolodihydropyridinones: 1,3-dipolar cycloaddition of nitrile oxides onto 2,4-dioxopiperidines

Abstract

Practical and efficient methods have been developed for the diversity-oriented synthesis of isoxazolodihydropyridinones via the 1,3-dipolar cycloaddition of nitrile oxides onto 2,4-dioxopiperidines. A select few of these isoxazolodihydropyridinones were further elaborated with triazoles by copper catalyzed azide-alkyne cycloaddition reactions. A total of 70 compounds and intermediates were synthesized and analyzed for drug likeness. Sixty-four of these novel compounds were submitted to the NIH Molecular Libraries Small Molecule Repository for high-throughput biological screening.

Introduction

Isoxazoles are an important class of nitrogen containing heterocycles found in many natural products and biologically active compounds.¹ While many methods have been developed for isoxazole formation, the cycloaddition of nitrile oxides to alkynes is the most effective.² Emphasis has been placed on the synthesis of highly substituted isoxazoles by Larock,³ Miyata,⁴ and others.⁵ While the reactions of nitrile oxides with cyclic 1,3-diketones⁶ and β-ketoesters⁷ have been studied, the cycloaddition of nitrile oxides onto heterocyclic 1,3-dicarbonyl compounds has, to our knowledge, not been reported. Although other routes have been reported for the preparation of isoxazolo[4,5-c]pyridinones,⁸ our method provides convergent access to a variety of diversity points around the core scaffold.

We recently reported the hydrolytic opening of the amide of a isoxazolodihydropyridinone substrate as a route to orthogonally protected heterocyclic diamino acids.⁹ As a continuation of this line of research, we report here the synthesis of a library of isoxazolodihydropyridinones via the 1,3-dipolar cycloaddition of nitrile oxides onto 2,4-dioxopiperidine to yield eleven novel 6,7-dihydroisoxazolo[4,5-c]2yridine-4(5H)-ones. Noting the many recent synthetic applications of copper catalyzed azide-alkyne cycloadditon (CuAAC) reactions for the rapid diversification of library collections (→ triazole diversification),¹⁰ we also report exploiting the CuAAC reaction to diversify a subset of these isoxazolodihydropyridinones leading to an additional 42 triazole-functionalized isoxazolo-dihydropyridinones. The resulting collection has been added to the NIH Molecular Libraries Small Molecule Repository for high-throughput biological screening.

Results and Discussion

As a starting point, Boc-protected 2,4-dioxopiperidines were synthesized from substituted β-amino acids by a three-step process that commenced with Boc protection of β-alanines (1) following a modified literature procedure (90–95% yield).¹¹ These N-protected amino acids (2) were then coupled to Meldrum’s acid using standard EDC and DMAP coupling conditions⁹. The resulting conjugates were immediately dissolved in EtOAc and refluxed for four hours to yield Boc-protected dioxopiperidines 3 in 85–90% yield over 2 steps (via the presumed intermediacy of a Meldrum’s acid-derived acylketene, which subsequently underwent intramolecular cyclization with the Boc-protected amine moiety; Scheme 1). These dioxopiperidines were then treated with NaH in THF followed by the slow addition of various N-hydroxybenzimidoyl chlorides 4{1–8} to give isoxazolodihydropyridinones 5{1–11} in 43–76% yield (Table 1). The requisite N-hydroxybenzimidoyl chlorides 4{1–8} were prepared from the corresponding benzaldehydes in 2 steps following well-established literature procedures.¹²

Scheme 1.

Synthesis of 2,4-dioxopiperidines 3 (R¹ = H and Me).

Table 1.

1,3-Dipolar cycloaddition of nitrile oxides onto 2,4-dioxopiperidines.

Entry	Compound	R1	R2	Yield
1	5{1}	H	2-MeOC6H4	52
2	5{2}	H	4-BrC6H4	64
3	5{3}	H	2-FC6H4	58
4	5{4}	H	4-NO2C6H4	54
5	5{5}	H	3-NO2C6H4	50
6	5{6}	H	2-NO2C6H4	68
7	5{7}	H	2-NO2,5-ClC6H3	50
8	5{8}	H	4-ClC6H4	55
9	5{9}	Me	2-MeOC6H4	76
10	5{10}	Me	2-NO2,5-ClC6H3	43
11	5{11}	Me	4-ClC6H4	64

A detailed NMR study of 5{3} was undertaken to verify its structure and completely assign all carbons and protons within the molecule. While many of the assignments could be made from the ¹H and ¹³C NMR spectra, we were unable to assign carbons C3, C4, C5 and C6 (see Figure 1) from these spectra alone. Therefore, several additional NMR experiments were undertaken, including INADEQUATE,¹³ HSQC,¹⁴ HMBC,¹⁵ and COSY.¹⁶ From the two-dimensional INADEQUATE experiment on 5{3} (Figure 1), we could assign all of its carbons, including C3, C4, C5 and C6. By analogy, the insights gained through this INADEQUATE proved useful in making complete carbon assignments in all of the isoxazolodihydropyridinones reported here; for example, we were able to completely assign carbons for alkyne derivative 6{1} (Figure 2). All of the spectra obtained for 5{3} are available in the Supporting Information.

Figure 1.

INADEQUATE-based assignments of C3, C4, C5 and C6 in 5{3}.

Figure 2.

Assignments for 6{1} based on the INADEQUATE data from 5{3}.

Three of the isoxazolodihydropyridinones delineated in Table 1 (5{1–3}) were selected for diversification through CuAAC reactions by first installing an alkyne moiety on the dihydropyridinone nitrogen. As outlined in Scheme 2, this chemistry commenced by first removing the Boc group with TFA in CHCl₃ (30 min) to yield the secondary amide in quantitative yield. Each amide was then deprotonated with NaH in dry THF and N-propargylated with propargyl bromide to yield the alkyne- containing scaffold 6{1–3) in 50–55% yield. Subsequent copper-mediated 1,3-dipolar cycloaddition to these substrates with eight azides (7{1–8}) delivered triazole products 8{1–3, 1–8} in good to excellent yields (Table 2). The azides employed in these CuACC reactions were generated in situ from the corresponding amines following one of two protocols: (i) alkyl 1°-amines were treated with the diazo-transfer reagent imidazole-SO₂N₃ to yield the azide as reported by Stick¹⁷ and (ii) aryl 1°-amines were treated with t-BuONO and subsequent ArN₂⁺ → ArN₃ conversion was effected with TMS-N₃ as reported by Moses.¹⁸ These in situ prepared azides, the alkynes (6{1–3}), CuSO₄, and sodium ascorbate were combined and the ensuing 1,3-dipolar cycloaddition reactions yielded triazole-substituted products 8{1–3,1–8} (Table 2).

Scheme 2.

CuAAC of the isoxazolodihydropyridinones.

Table 2.

CuAAC diversification of isoxazolodihydropyridinones.

Entry	Azide	8{1,1–8} % yield	8{2,1–8} % yield	8{3,1–8} % yield
1	{1}	28	85	89
2	{2}	571	95	92
3	{3}	52	76	79
4	{4}	46	92	88
5	{5}	76	74	76
6	{6}	78	74	94
7	{7}	84	79	77
8	{8}	70	89	92

¹Decomposed upon isolation.

Two additional isoxazolodihydropyridinones delineated in Table 1 (5{4–5}) were selected for diversification through CuAAC reactions by taking advantage of the nitro group for azide entry. Isoxolopyridinones 5{4–5} were treated with Zn/AcOH to deliver the m- and p-anilines 9{1–2} in quantitative yields (Scheme 3). These anilines were then treated with ^tBuONO and the targeted azides were obtained by displacement of the diazonium salt with TMS-N₃ to yield 10{1–2} in 45% and 87% yield (meta and para), respectively. The Boc group in 10{1–2} was then removed by TFA treatment in CHCl₃ to give 11{1–2} in quantitative yield. At this juncture, N-alkylation of the 2°-amide in 11{1–2} proceeded smoothly by deprotonation with NaH in dry THF and subsequent treatment with 2-bromo-1-morpholinoethanone (12) to give 13{1–2} in 64% and 69% yield, respectively. 2-Bromo-1- morpholinoethanone 12 was prepared by dropwise addition of morpholine to a solution of bromoacetyl bromide in dry DCM under nitrogen (clear oil; 82% yield).¹⁹ Nine alkynes (Scheme 3) were then used to diversify this azide set to give 16 additional triazole products (15{1–2, 1–9}; Table 3).

Scheme 3.

Azide synthesis for CuAAC.

Table 3.

Diversification of the alkylated isoxazolodihydropyridinones.

Entry	Alkyne	m-N3 yield	p-N3 yield
1	{1}	28	79
2	{2}	69	47
3	{3}	51	99
4	{4}	89	99
5	{5}	36	32
6	{6}	57	93
7	{7}	68	99
8	{8}	47	65
9	{9}	85	62

A Lipinski rule-of-five analysis²⁰ was performed to evaluate the drug likeness of all compounds and intermediates reported (Figure 3). Of the seventy compounds analyzed, fifty-eight were within the parameters set by Lipinski. Of the remaining, nine from chemset 15 had >10 hydrogen bond acceptors and the remaining two compounds (also from chemset 15) had two violations each (>10 acceptors and a mass >500 Daltons; both had molecular weights of ~560 g/mol). The other chemsets combined had only one violation (e.g., 8{2,8}), suggesting that this collection of compounds is very appropriate for high-throughput screening to discover drug leads and biological probes.

Figure 3.

Lipinski rules analysis for 70 compounds and intermediates; calculated using Molinspirations Online Molecular Properties Calculator [http://www.molinspiration.com/cgi-bin/properties?textMode=1].

Conclusions

A short, reliable, and efficient convergent method has been developed to produce a diverse array of novel isoxazolodihydropyridinones starting from commercially available beta-amino acids and benzaldehydes. Eleven isoxazolpyridinones were synthesized and five of these were diversified further with CuAAC chemistry to give 42 additional isoxazolodihydropyridinone-triazole products. In all, 64 of the 70 compounds and intermediates synthesized were acceptable for submission to the NIH Molecular Libraries Small Molecule Repository for high-throughput biological screening; the unsubmitted 6 compounds contained excluded aromatic azides.