A one-step, atom economical synthesis of thieno[2,3-d]pyrimidin-4-amine derivatives via a four-component reaction

Abstract

A Na₂HPO₄-catalyzed four-component reaction between a ketone, malononitrile, S₈ and formamide has been realized for the first time. This reaction provides a concise approach to thieno[2,3-d]pyrimidin-4-amines, previously requiring 5 steps. The utility of this reaction was validated by preparing a multi-targeted kinase inhibitor and an inhibitor of the NRF2 pathway with excellent atom- and step-economy.

Graphical Abstract

Concise synthesis of thieno[2,3-d]pyrimidin-4-amines with improved atom economy, increased step economy, enhanced structural diversity and simplified purification has been realized.

Introduction

Thieno[2,3-d]pyrimidin-4-amines (TPAs) have been discovered to possess many important bioactivities, including inhibition of hepatitis C virus (HCV) replication^[1], inhibition of kinases,^[2] modulation of the NRF2 anti-oxidant response pathway,^[3]and inhibition of human farnesyl pyrophosphate synthase (hFPPS)^[4] (Figure 1). Notably, GDC-0941 is currently being evaluated as an anticancer agent in a phase II clinical trial,^[2a] while GDC-0980 and SNS-314 are both being tested as anticancer agents in phase I clinical trials^[2b],[2c]. In light of the biological and pharmaceutical value of TPAs, the organic synthesis community has been interested in facile syntheses of these molecules. Currently, the reported synthesis of TPAs requires 5 steps starting with malononitrile and acetyltrimethylsilane (Scheme 1, top).^[4] To develop a general, straightforward method of synthesizing TPAs, we designed a one-step synthesis using a four-component reaction between ketones, malononitrile, S₈ and formamide based on the Gewald reaction (Scheme 1, bottom).^[5] This design is expected to improve atom economy, ^[6] step economy,^[7] structural diversity^[8] and ease of purification.

Figure 1.

Bioactive molecules bearing a thieno[2,3-d]pyrimidin-4-amine core.

Scheme 1.

The previously reported route of TPA synthesis and the presently reported route.

Results and Discussion

To test our proposed atom-economical thieno[2,3-d]pyrimidin-4-amine synthesis, we started with 1 equivalent of acetophenone (1a), 1.5 equivalents of malononitrile (2), 1.1 equivalents of S₈ (3) and 12 equivalents of formamide (4). We were surprised to find that by simply heating the reaction to 170°C, even in the absence of a catalyst, we could isolate the desired product 5a in 34% yield (Table 1, Entry 1). Next, we evaluated a panel of potential catalysts including L-proline/Et₂NH, Na₂CO₃, Na₂HPO₄, K₂CO₃, and K₃PO₄ (Table 1, Entries 2–6), and found Na₂HPO₄ as the optimal catalyst giving 5a in 85% yield (Table 1, Entry 4). Compared to other catalysts tested, Na₂HPO₄ is the only one that could function as both a base and an acid, which is a possible reason for its superior catalytic effects. Using Na₂HPO₄ as the catalyst, we next explored solvent effects (Table 1, Entries 7–10). Water, tert-butanol, sec-butanol, and isopropanol all caused a decrease in the yield of 5a. This result indicated that the reactant formamide, 4, is also an excellent solvent for the reaction. Next, in order to optimize the efficiency of the reaction, we explored the minimum amount of 4 that could be used (Table 1, Entries 4, 11–13). With 4 eq. of 4, the yield of 5a was slightly lower, but when we investigated the substrate scope with 4 eq. of 4, we found many of the substrates gave a poor yield in these conditions. When the amount of 4 dropped to 2 eq., the yield of 5a decreased from 86% to 75%, and was further reduced to 62% if 1.5 eq. of 4 was used. Therefore, 12 eq. of 4 gives the highest yield and, importantly, the greatest substrate scope. We next explored time and temperature and found the yield increased to 91% and the reaction time could be decreased from 6 hours to 0.5 hour by increasing the temperature to 200°C (Table 1, Entry 14). Further, in the presence of 10 mol% triphenylphosphine as an additive, the yield of 5a further increased to 96% (Table 1, Entry 15). But with Ph₃P and no Na₂HPO₄, the yield dropped to 55% (Table 1, Entry 16). The presence of Ph₃P possibly prevents the dimerization of 5a.^[9] Overall, we chose 20 mol% Na₂HPO₄ as catalyst, 10 mol% Ph₃P, and 200°C for 0.5 h, as the standard conditions for a survey of the substrate scope.

Table 1.

Optimization of conditions of a four-component reaction between acetophenone 1a, malononitrile 2, S₈ 3, and formamide 4.^[a]

Entry	Catalyst	Eq. of 4	Solvent	T/°C	% Yield[d]
1	None	12	-	170	34
2	L-proline/Et2NH	12	-	170	61
3	Na2CO3	12	-	170	35
4	Na2HPO4	12	-	170	85
5	K2CO3	12	-	170	52
6	K3PO4	12	-	170	45
7	Na2HPO4	12	H2O	170	10
8	Na2HPO4	12	t-BuOH	170	45
9	Na2HPO4	12	s-BuOH	170	41
10	Na2HPO4	12	i-PrOH	170	32
11	Na2HPO4	4	-	170	84
12	Na2HPO4	2	-	170	75
13	Na2HPO4	1.5	-	170	62
14[b]	Na2HPO4	12	-	200	91
15[b], [c]	Na2HPO4	12	-	200	96
16[b]	PPh3	12	-	200	55

^[a]1.0 eq. ketone (1a), 1.5 eq. malononitrile (2), 1.1 eq. S₈ (3), 12.0 eq. formamide (4), and 20 mol% catalyst were used for 6 hours.

^[b]0.5 hour.

^[c]10 mol% Ph₃P added.

^[d]Isolated yields.

After optimizing conditions, a set of aromatic moieties at the R¹ position were evaluated. Electron-withdrawing groups (EWG) gave an 81% average yield (Scheme 2, 5b and 5c). Para-halide substituents gave an 85% average yield (Scheme 2, 5d-5g). Meta- and ortho-F substituents gave 85% and 82% yields, respectively. Electron-donating groups (EDG) offered desired products in 77% average yield (Scheme 2, 5j-5m). The structure of 5l was unambiguously assigned by X-ray crystallography (Figure 2; CCDC 1904050). Heterocyclic groups including furanyl and thiophenyl generated 5n and 5o in 62% and 65% yields, respectively.

Scheme 2.

Substrate scope of a four-component reaction of ketones 1, malononitrile 2, S₈ 3, and formamide 4.

Figure 2.

ORTEP diagram of 5I with atomic labeling scheme. Thermal ellipsoids at a 50 % probability level are shown. CCDC 1904050.

Next, we set R¹ = Phenyl, and looked at the effect of changing the R² group. Both phenyl and methyl groups had excellent compatibility in the reaction, providing 5p and 5q with 97% and 91% yields, respectively. Finally, an alkyl group at both R¹ and R² was tested, producing 5r in 87% yield. Generally, the reaction has broad ketone substrate scope.

Scalability of an organic reaction is one of the important features when it is used by medicinal chemists and chemical process scientists.^[10] We were able to scale-up the reaction to 50 mmol, furnishing 9 grams of product 5a in 79% yield after a simple work-up and recrystallization (Scheme 3).

Scheme 3.

Scalable four-component reaction of acetophenone 1a, malononitrile 2, S₈ 3 and formamide 4.

The practicality of the reaction was demonstrated by the rapid synthesis of the NRF2 regulator AEM1 and the multi-target kinase inhibitor (MTKI; Scheme 4). AEM1 could be obtained via a conventional N-alkylation of 5c with 74% overall yield in two steps from the readily available 1c. This is the first disclosure of the synthesis of AEM1. MTKI was synthesized with a 49% overall yield in 2 steps from the inexpensive para-nitro starting material, 1i. The quick access to intermediate 5i via the four-component reaction of 1i, 2, 3, and 4 makes this route an attractive alternative to the preparation the MTKI.

Scheme 4.

Total synthesis of AEM2 and MTKI using a Na₂HPO₄-catalyzed four-component reaction of ketones (1c or 1i), malononitrile (2), S₈ (3) and formamide (4).

Finally, we propose a possible mechanism of the four-component reaction of ketones, malononitrile, S₈ and formamide (Scheme 5). The reaction is initiated by Na₂HPO₄ extracting the proton of 2. Then the ketone is subjected to nucleophilic attack of activated 2, leading to the formation of intermediate A. The product of a Knoevenagel reaction^[11], B, is generated through the elimination of A with the help of Na₂HPO₄. Subsequently, Na₂HPO₄ activated B opens the S₈ ring through nucleophilic attack to yield intermediate C. Intermediate C undergoes intramolecular addition to the cyano group to form D. Then Na₂HPO₄-promoted aromatization of D leads to the formation of Gewald reaction product E.^[12] Condensation of E and 4 produces intermediate F which undergoes intramolecular addition to the cyano group to give G. Finally, the Na₂HPO₄-catalyzed tautomerization of G gives the product 5.

Scheme 5.

Proposed mechanism of the Na₂HPO₄-catalyzed four-component reaction of ketone, malononitrile, S₈ and formamide

Conclusion

In conclusion, we have developed a four-component reaction between ketones, malononitrile, S₈ and formamide, to prepare TPAs in 62–97% yield. This reaction provides a facile synthesis of high value intermediates with excellent atom economy, short reaction times, high structural diversity, and facile purification. In addition, the reaction could be scaled up to a multi-gram scale that did not require chromatography for purification. To demonstrate the general applicability of our four-component reaction, the reaction was applied to prepare the NRF2 regulator AEM2 and MTKI in two steps with high overall yields. In the future, the invention of this reaction is expected to accelerate the synthesis of TPA-based molecules for both drug discovery efforts and process chemistry.

Experimental Section

General procedure for the preparation of compounds 5a-r.

To a 10 mL reaction vial, was added 1.0 mmol (1.0 eq.) ketone (1a-r), 1.5 mmol (1.5 eq.) malononitrile (2), 1.1 mmol (1.1 eq.) S₈ (3), 12 mmol (12.0 eq.) formamide (4), 20 mol% Na₂HPO₄, and 10 mol% PPh₃. The mixture was heated to 200°C in a sand bath for 0.5 hour. 1 was confirmed to be consumed by TLC analysis. Then 2 mL of water was poured into the reaction mixture after it cooled to 22°C. The resulting mixture was extracted with ethyl acetate (EtOAc) 3 times (5 mL EtOAc per extraction). The organic layers were combined and washed with 15 mL water. Subsequently, the organic layer was dried with anhydrous magnesium sulfate. The crude product was obtained after removing the EtOAc with vacuum rotation. Further purification was performed with flash column chromatography to afford products 5a-r as solids.