Efficient Synthesis of 4-Amino-2-methoxy-7,8-dihydropyrido[4,3-d]pyrimidin-5-ones: Practical Access to a Novel Chemotype in the Development of DGAT-1 Inhibitors

Abstract

A practical access to an unprecedented, fused bicyclic 4-amino-2-alkoxy-7,8-dihydropyrido[4,3-d]pyrimidin-5-one scaffold is developed. The synthesis of the potent inhibitor, 2-{4-[4-amino-2-methoxy-5-oxo-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl]phenyl}-2-methylpropanamide is detailed, with particular emphasis placed on synthetic efficiency and scalability. With the isolation of solid intermediates, the routes described offer clear elements of practicality and facilitate production of the target compound on large scale (>10 g) without chromatography.

Introduction

Inhibition of acyl-CoA:diacylglycerol acyltransferase-1 (DGAT-1), the enzyme which catalyzes the rate-limiting step in triglyceride synthesis, holds promise for the treatment of Type II diabetes mellitus (T2DM) and obesity.[ ¹ ] As part of a program directed toward the identification of orally active DGAT-1 inhibitors, our team recently reported the discovery of PF-04620110 (1) (Scheme [1]), which is currently in clinical trials for the treatment of T2DM.[ ² ] Following the discovery of this carboxylic acid based clinical candidate, a back-up program was launched to identify inhibitors based on a neutral chemotype. These efforts led to the discovery of a DGAT-1 inhibitor series based on a novel heterocyclic scaffold, containing a 4-amino-2-methoxy-7,8-dihydropyrido[4,3-d]pyrimidin-5-one scaffold.[ ³ ] Based on promising in vitro and in vivo profiles, orally bioavailable amide 2 was selected for in-depth preclinical evaluation. This article showcases the development and optimization of synthetic routes to this DGAT-1 inhibitor and its novel ring system.

A retrosynthetic analysis of DGAT-1 inhibitor 2 is depicted in Scheme [2]. We hypothesized that the aminopyrimidine motif found in N-aryllactam template 3 could arise from bis-amination of 3-cyano-4-hydroxydihydropyridone 4. In turn, scission of key intermediate 4 at the amide and double bonds revealed N-monoalkylated aniline 5 as a readily available precursor. In practice, two disconnections of aniline 5 were explored: a Buchwald amination[ ⁴ ] of an appropriate aryl halide with a β-alanine ester (route A), and an N-alkylation of the corresponding aniline via an aza-Michael addition with an acrylate (route B). Typically, anilines 5 were not isolated, and purification was performed instead at the vinylogous carbamate stage (intermediate 4). Overall yields between the two approaches were quite comparable, and in practice, route selection toward analogues in this novel series was guided by reaction impurity profiles, reproducibility, and the compatibility of functional groups with reaction conditions. In the case of amide 2, both of these attractive approaches were executed and are detailed herein.

Results and Discussion

The synthesis of DGAT-1 inhibitor 2, according to the retrosynthetic plan depicted in Scheme [2], required access to vinylogous carbamate 12, and two successful approaches to this key intermediate are outlined in Scheme [3]. Commencing with route A, commercially available ester 6 was subjected to α,α-methylation to provide bromide 7. Palladium-catalyzed Buchwald amination of bromide 7 with β-alanine ethyl ester hydrochloride produced aniline 8, using 2-dicyclohexylphosphino-2′,4′,6′-triisopropylbiphenyl (X-Phos)[ ⁵ ] and cesium carbonate as the optimal combination of ligand and base. We observed that the reaction went to completion more efficiently above 100 °C, and with the addition of N,N-diisopropylethylamine (DIPEA). Excess amine was used to suppress the formation of by-product 9, which arises from N-arylation of the desired aniline 8.[ ⁵ ] Though separable by chromatography, crude mixtures of 8 and 9 were typically carried onto the cyanoacetylation (step c) and Dieckmann-type cyclocondensation (step d)[ ⁶ ] to afford, after aqueous work-up, pure hydroxydihydropyridone 12. Indeed, only aniline 8 underwent N-acylation to afford cyclocondensation precursor 10. Treatment of a mixture of nitrile 10 and amination by-product 9 with 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) in warm methanol induced a mild condensation to provide vinylogous carbamate salt 11.[ ⁷ ] The high solubility of 1,8-diazabicyclo[5.4.0]undec-7-ene salt 11 in water was key in facilitating the extractive removal of lipophilic impurities, such as bis-N-arylamine 9 and residual ligand from the amination step. In the optimized Dieckmann-type process (step d), the methanol was removed upon consumption of 10, and the residue was partitioned between water and ethyl acetate. The basic aqueous layer was then acidified to induce the precipitation of pure vinylogous carbamate 12. Thus, the four-step process following route A from bromide 6 delivered routinely tens of grams of key substrate 12 in overall yields exceeding 30%, using simple extraction protocols and without the need for chromatography.

Route B, featuring an aza-Michael addition, also proved effective in generating aniline 8 en route to key vinylogous carbamate 12 (Scheme [3]). Mild bis-alkylation of commercially available methyl 2-(4-nitrophenyl)acetate (13) afforded ester 14 in 79% isolated yield. Hydrogenation with palladium-on-carbon in methanol then provided known aniline 15 [ ⁸ ] in nearly quantitative yield. Finally, N-alkylation of aniline 15 with ethyl acrylate provided aniline 8 as a mixture contaminated with small amounts of over-alkylation product 16 and unconsumed starting material 15.[ ⁹ ] Attempts to drive the aza-Michael addition to completion only increased the production of the undesired bis-N-alkylaniline 16. Inconsequentially however, this mixture was carried forward through N-acylation and cyclization, as described for route A, to permit the isolation of pure carbamate 12 without chromatography. Route B thus provided pure vinylogous carbamate 12 in an overall yield of 58% over the five steps.

Key carbamate 12 was now primed for a high-yielding, five-step sequence to forge the 4-amino-2-methoxypyrimidine domain found in 2 (Scheme [4]). This de novo pyrimidine construction was initiated by chlorination with oxalyl chloride (COCl)₂, followed by methanolysis to generate vinylogous methylcarbamate 17 in high yield and purity. While O-methylation of 12 was possible with trimethylsilyldiazomethane,[ ¹⁰ ] the chlorination–methoxylation protocol demonstrated a cleaner reaction profile and was especially preferred on larger scales. Next, the addition of sodium methoxide to a methanolic slurry of cyanamide and electrophile 17 provided adduct 18 in nearly quantitative yield. Construction of the aminopyrimidine moiety was then accomplished by acid-catalyzed cyclization in methanol[ ¹¹ ] to provide ester 19 as a key intermediate, albeit as a mixture contaminated with amine 20 as a major impurity (ca. 10%). It remains unclear whether by-product amine 20 arises from solvolysis of the cyanamide moiety of 18, or from Michael-type substitution by an opportunistic ammonium species.[ ¹² ] Notably, the two-step sequence from 17 into pyrimidine 19 could be performed in a single flask, without the isolation of cyanamide 18 or significant loss in yield, by simple acidification and heating of the reaction mixture from step b. Methoxypyrimidine 19 was also available directly from vinylogous carbamate 17 upon addition of O-alkylisoureas (step f),[ ¹³ ] however, these direct transformations typically gave increased amounts of by-products such as 20, and often required rigorous chromatographic separations to isolate the desired targets. Hydrolysis of methyl ester 19 into carboxylic acid 21 was not straightforward; we observed that the harsh conditions required to saponify the hindered ester also hydrolyzed the labile pyrimidinyl methoxy group. Extensive optimization of this step revealed that potassium trimethylsilanolate (KOTMS) was more effective in driving the chemoselective hydrolysis of the ester to completion without side reactions. Carboxylic acid 21 was then activated with 1,1′-carbonyldiimidazole (CDI), and mild aminolysis with aqueous ammonium hydroxide finally completed the synthesis to deliver DGAT-1 inhibitor 2 in high yield and purity, after recrystallization from acetic acid–water. The overall yield of the sequence from intermediate 12 to amide 2 (steps a–e, Scheme [4]) was 57% over the five steps. The optimized routes described in Schemes 3 and 4 enabled the advancement of commercially available acetate 13 into target 2 in 33% overall yield over ten operations without the need for chromatographic purifications.

Summary

In conclusion, we have described robust and reliable synthetic routes to an unprecedented 4-amino-2-methoxy-7,8-dihydropyrido[4,3-d]pyrimidin-5-one bicyclic ring system, which was developed to enable the identification of potent, orally bioavailable DGAT-1 inhibitors. The utility of two versatile approaches into this ring system (routes A and B, Scheme [3]) were demonstrated with the scalable synthesis of DGAT-1 inhibitor 2 in high overall yield. The synthetic steps described were performed on scales of ten to one hundred grams, and the materials delivered by these sequences permitted the rapid in vitro and in vivo profiling required for preclinical evaluation. Notably, the development of practical acid–base washes and recrystallizations enabled the easy isolation of solid intermediates in high purity and obviated completely the need for column chromatography. The processes described herein should prove useful to synthetic chemists interested in similar organic transformations or in the efficient synthesis of related structural motifs.

All reagents and solvents were of commercial quality and used as received, unless otherwise noted. Silica gel (40 μm) was purchased from J. T. Baker, Phillipsburg, NJ (USA). Melting points were recorded on a Thomas Hoover MeltTemp apparatus and are uncorrected. Proton (400 or 500 MHz) and carbon (100 or 125 MHz) NMR spectra were recorded at room temperature on Varian Unity™ 400 or 500 spectrometers. Chemical shifts are expressed in parts per million (δ) relative to residual solvent as an internal reference. The signals are denoted as follows: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br s, broad singlet. Mass spectrometry was performed by direct flow analysis using positive and negative atmospheric pressure chemical ionization (APCI) or electron scatter (ES) ionization sources. Liquid chromatography–mass spectrometry (LC–MS) was performed on an Agilent 1100 Series [Waters Atlantis C18 column, 4.6 × 50 mm, 5 μm; 95% H₂O–MeCN linear gradient to 5% H₂O–MeCN over 4 min, hold at 5% H₂O–MeCN to 5 min, TFA modifier (0.05%); flow rate = 2.0 mL/min]. Gas chromatography–mass spectrometry (GC–MS) was performed on a HP6890 Series GC System (Agilent 19091 capillary column, 0.2 × 12 m, 0.33 μm; 10.5 psi; flow rate = 40 mL/min; temperature = 105–250 °C). High-resolution mass spectrometry (HRMS) was performed on an Agilent (6620) LC–MS TOF using an Xbridge C18 3.0 × 5.0 mm, 2.5 μm column at 60 °C; NH₄HCO₂–H₂O (1:99, w/w) as mobile phase A1 and MeOH–MeCN (50:50) as mobile phase B2. A Waters APCI/MS model ZMD mass spectrometer, or a Waters/Micromass ESI/MS model ZMD or LCZ mass spectrometer, each equipped with a Gilson 215 liquid handling system and HP 1100 DAD, were used to carry out the experiments. Where the intensities of chlorine or bromine-containing ions are described, the expected intensity ratio was observed (approximately 3:1 for ³⁵Cl/³⁷Cl-containing ions and 1:1 for ⁷⁹Br/⁸¹Br containing ions) and only the lower mass ion is given. Unless otherwise noted, mass ion peaks are reported for all examples.

Methyl 2-(4-Bromophenyl)-2-methylpropanoate (7)

To a soln of methyl 2-(4-bromophenyl)acetate (6) (203 g, 0.89 mol) and MeI (378 g, 2.66 mol) in THF (1.5 L) was added portionwise, under Ar, 60% NaH in mineral oil (88.6 g, 2.22 mol), with external cooling such that the internal temperature was maintained at 45–50 °C. The mixture was stirred for 3 h at 50 °C and then poured into H₂O (3 L). The product was extracted into MTBE (2 × 1.25 L). The combined organic layer was washed with H₂O (2 L) and brine (500 mL), and concentrated to a pale-yellow oil. Vacuum distillation (120–123 °C, 5 Torr) of this oil gave the desired ester 7.

Yield: 151.2 g (66%); colorless oil.

¹H NMR (400 MHz, CDCl₃): δ = 7.42 (d, J = 8.7 Hz, 2 H), 7.20 (d, J = 8.8 Hz, 2 H), 3.63 (s, 3 H), 1.54 (s, 6 H).

¹³C NMR (100 MHz, CDCl₃): δ = 176.9, 143.9, 131.7, 127.8, 120.9, 52.5, 46.5, 26.7.

HRMS (ESI): m/z [M + H]⁺ calcd for C₁₁H₁₄BrO₂: 257.0177; found: 257.0175.

Methyl 2-{4-[3-Cyano-4-hydroxy-2-oxo-5,6-dihydropyridin-1(2H)-yl]phenyl}-2-methylpropanoate (12); Route A

Pd(OAc)₂ (1.53 g, 6.81 mmol, 2.5 mol%) and X-Phos (6.48 g, 13.6 mmol, 5 mol%) were added to toluene (2.5 L), and the mixture was degassed with Ar over 30 min. Bromide 7 (70.0 g, 272 mmol) and Cs₂CO₃ (265 g, 817 mmol, 3.0 equiv) were added, and the resulting mixture was degassed with Ar for 10 min. β-alanine ethyl ester hydrochloride (83.6 g, 544 mmol, 2.0 equiv) and DIPEA (83.6 g, 645 mmol, 2.29 equiv) were added and the mixture heated at 120 °C for 36 h. The mixture was cooled to r.t. and filtered through a pad of Celite. The solvent was removed in vacuo to afford aniline 8 as an oil (108 g) contaminated with by-product 9 (ca. 10% by ¹H NMR spectroscopy). This crude mixture was used without further purification. An analytical sample of 8 was isolated by chromatography.

¹H NMR (400 MHz, CDCl₃): δ = 7.16 (d, J = 8.8 Hz, 2 H), 6.59 (d, J = 8.8 Hz, 2 H), 4.15 (q, J = 6.7 Hz, 2 H), 4.14 (br s, 1 H), 3.62 (s, 3 H), 3.42 (t, J = 6.5 Hz, 2 H), 2.59 (t, J = 6.5 Hz, 2 H), 1.53 (s, 6 H), 1.25 (t, J = 6.7 Hz, 3 H).

Crude aniline 8 from the previous step (theoretically 272 mmol) was dissolved in CH₂Cl₂ (1.2 L), and cyanoacetic acid (46.5 g, 544 mmol, 2.0 equiv), DMAP (1.7 g, 5 mol%) and EDC (73.3 g, 381 mmol, 1.4 equiv) were added. The mixture was stirred at r.t. for 16 h then concentrated to dryness. The residue was partitioned between EtOAc (1 L) and H₂O (1 L). The organic layer was separated, washed sequentially with sat. aq NaHCO₃ soln (2 × 500 mL), sat. aq NH₄Cl soln (250 mL), and brine (250 mL), then dried over Na₂SO₄. The soln was filtered and concentrated to a wax containing amide 10, which was used in the next step without further purification. An analytical sample of 10 was isolated by chromatography.

¹H NMR (400 MHz, CDCl₃): δ = 7.42 (d, J = 8.8 Hz, 2 H), 7.16 (d, J = 8.8 Hz, 2 H), 4.12–3.95 (m, 4 H), 3.67 (s, 3 H), 3.15 (s, 2 H), 2.59 (t, J = 6.5 Hz, 2 H), 1.57 (s, 6 H), 1.25 (t, J = 6.7 Hz, 3 H).

Crude amide 10 from the previous step (theoretically 272 mmol) was dissolved in MeOH (1 L). DBU (48.9 mL, 327 mmol, 1.2 equiv) was added, and the mixture heated at 50 °C for 10 h. The solvent was removed in vacuo and replaced with H₂O (500 mL). EtOAc­ (1 L) was added and the mixture stirred vigorously for 15 min. The basic aq layer was removed and washed with EtOAc–heptane (1:1, 500 mL). The aq layer was then acidified to ca. pH 2 by the dropwise addition of aq HCl (3 M). Precipitation was observed, and upon completion of the addition, the mixture was stirred for 15 min at r.t. The solid was filtered and rinsed sequentially with H₂O (2 × 500 mL), EtOAc–heptane (1:1, 250 mL) and heptane (250 mL), and then dried in vacuo at 45 °C to afford vinylogous carbamate 12.

Yield: 41.9 g (49% over three steps from bromide 7); off-white powder; mp 238–241 °C.

¹H NMR (400 MHz, DMSO-d ₆): δ = 7.29–7.24 (m, 2 H), 7.21–7.17 (m, 2 H), 3.75 (t, J = 6.8 Hz, 2 H), 3.56 (s, 3 H), 2.77 (t, J = 6.7 Hz, 2 H), 1.47 (s, 6 H).

¹³C NMR (100 MHz, DMSO-d ₆): δ = 182.2, 177.0, 163.1, 142.3, 141.5, 126.4, 125.7, 115.6, 84.3, 52.7, 46.5, 46.2, 29.8, 27.0.

HRMS (ESI): m/z [M + H]⁺ calcd for C₁₇H₁₉N₂O₄: 315.1345; found: 315.1345.

Methyl 2-Methyl-2-(4-nitrophenyl)propanoate (14)

To a soln of methyl 4-nitrophenylacetate (13) (100 g, 0.51 mol) and MeI (128 mL, 2.05 mol), 4.0 equiv) in THF (500 mL) and DMF (100 mL) was added dropwise a soln of t-BuOK (1.53 L, 1.53 mol, 1 M, 3.0 equiv) in THF. The reaction temperature was maintained at 50 °C during the addition, after which the mixture was stirred at 50 °C for 1 h. The mixture was cooled to r.t. and poured into HCl (1 L, 2 M) with stirring. Toluene (1 L) was added, and the organic layer was washed with H₂O (500 mL) and brine (500 mL), and then dried (MgSO₄), filtered and evaporated to afford a red oil. Hexane (200 mL) was added, and upon standing at r.t. overnight the product crystallized and was filtered and dried to provide ester 14.

Yield: 90 g (79%); red solid; mp 45–46 °C.

¹H NMR (400 MHz, CDCl₃): δ = 8.15 (d, J = 8.8 Hz, 2 H), 7.50 (d, J = 8.8 Hz, 2 H), 3.67 (s, 3 H), 1.61 (s, 6 H).

¹³C NMR (100 MHz, CDCl₃): δ = 176.2, 152.2, 146.9, 127.1, 123.7, 52.7, 47.1, 26.6.

MS (ESI): m/z = 224.0 [M + H]⁺.

HRMS (ESI): m/z [M + H]⁺ calcd for C₁₁H₁₄NO₄: 224.0923; found: 224.0925.

Methyl 2-(4-Aminophenyl)-2-methylpropanoate (15)

A mixture of nitroarene 14 (125.0 g, 560 mmol) and Pd/C (5 g, 5% w/w, wet, 0.5 mol%) in MeOH (550 mL) was subjected to an atmosphere of H₂ (50 psi) at r.t. for 20 h. The mixture was filtered through Celite and concentrated to provide aniline 15, which was used in the next step without further purification.

Yield: 103.9 g (96%); red-orange wax.

¹H NMR (400 MHz, CDCl₃): δ = 7.12 (d, J = 8.8 Hz, 2 H), 6.65 (d, J = 8.8 Hz, 2 H), 3.61 (s, 3 H), 1.53 (s, 6 H).

¹³C NMR (100 MHz, CDCl₃): δ = 177.9, 145.2, 134.9, 126.8, 115.2, 52.3, 45.9, 26.8.

HRMS (ESI): m/z [M + H]⁺ calcd for C₁₁H₁₆NO₂: 194.1181; found: 194.1180.

Methyl 2-{4-[3-Cyano-4-hydroxy-2-oxo-5,6-dihydropyridin-1(2H)-yl]phenyl}-2-methylpropanoate (12); Route B

Aniline 15 (93.5 g, 484 mmol) was dissolved in ethyl acrylate (58 mL, 532 mmol, 1.1 equiv) and glacial AcOH (32 mL, 532 mmol, 1.1 equiv). The mixture was heated at 50 °C with mechanical stirring for 12 h, then cooled to r.t., and diluted with toluene (250 mL) and 10% aq K₂CO₃ soln (125 mL). After stirring for 1 h, the organic layer was washed with brine (125 mL), dried over Na₂SO₄, filtered and concentrated to a dark brown oil, which solidified under vacuum to afford Michael adduct 8 as an amber wax (154 g). This crude mixture was used in the next step without further purification.

Crude alkylated aniline 8 from the previous step (theoretically 484 mmol) was dissolved in EtOAc (1.35 L). Cyanoacetic acid (41.2 g, 484 mmol, 1.0 equiv) and Et₃N (196 mL, 1.40 mol, 2.9 equiv) were then added sequentially, and the mixture cooled to 0 °C. A 50% soln of 1-propanephosphonic cyclic anhydride (T3P) in EtOAc (308 mL, 484 mmol, 1.0 equiv) was added dropwise over 20 min, at a rate such that the internal reaction temperature did not exceed 10 °C. After warming to r.t. over 30 min, the mixture was diluted with EtOAc (400 mL) and washed sequentially with 10% aq K₃PO₄ soln (900 mL), HCl (1.8 L, 1 M), and brine (900 mL). The organic layer was dried over Na₂SO₄ and concentrated to provide N-acylaniline 10 as a pale yellow wax (204 g). This crude material was used in the next step without further purification.

Crude amide 10 from the previous step (theoretically 983 mmol) was dissolved in MeOH (2 L). DBU (176 mL, 1180 mmol, 1.2 equiv) was added, and the mixture heated at 70 °C for 2 h. The solvent was removed in vacuo and the residue partitioned between EtOAc (750 mL) and HCl (750 mL, 1 M) with stirring. Heptane (750 mL) was slowly added over 30 min, inducing the precipitation of a well-dispersed solid. After stirring for 30 min at r.t., the mixture was filtered, and the filter cake washed with H₂O (300 mL) and heptane–EtOAc (1:1, 300 mL), and dried under vacuum at 50 °C to afford vinylogous carbamate 12 as an off-white powder (117 g, 77% over 3 steps). The analytical data matched those reported above for 12 prepared via route A.

Methyl 2-{4-[3-Cyano-4-methoxy-2-oxo-5,6-dihydropyridin-1(2H)-yl]phenyl}-2-methylpropanoate (17)

A flask was charged with vinylogous carbamate 12 (50.0 g, 159 mmol), catalytic DMF (100 μL), and CH₂Cl₂ (625 mL) at 0 °C. (COCl)₂ (25.5 mL, 294 mmol, 1.85 equiv) was added dropwise over 20 min, and the mixture warmed to r.t. over 1 h, at which point LC–MS indicated the complete formation of the vinylogous chloride intermediate. MeOH (725 mL) was then added via an addition funnel at r.t., and the resulting mixture heated at reflux temperature overnight. The CH₂Cl₂ was removed under reduced pressure to afford a thick slurry, which was filtered. The white solid collected was washed with MeOH (250 mL) and dried under vacuum at 50 °C to afford vinylogous ester 17.

Yield: 45.5 g (87%); off-white powder; mp 191–193 °C.

¹H NMR (400 MHz, DMSO-d ₆): δ = 7.30–7.26 (m, 2 H), 7.25–7.20 (m, 2 H), 4.01 (s, 3 H), 3.82 (t, J = 6.8 Hz, 2 H), 3.56 (s, 3 H), 3.02 (t, J = 6.8 Hz, 2 H), 1.47 (s, 6 H).

¹³C NMR (100 MHz, DMSO-d ₆): δ = 183.3, 176.9, 162.2, 142.6, 141.2, 126.5, 125.7, 114.9, 86.6, 58.5, 52.8, 46.5, 46.0, 27.1, 25.9.

HRMS (ESI): m/z [M + H]⁺ calcd for C₁₈H₂₁N₂O₄: 329.1501; found: 329.1502.

Methyl 2-{4-[4-Amino-2-methoxy-5-oxo-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl]phenyl}-2-methylpropanoate (19)

The vinylogous ester 17 (15.0 g, 45.7 mmol) was suspended in MeOH (150 mL) at 0 °C, after which cyanamide (4.20 g, 100.5 mmol, 2.2 equiv) was added. NaOMe (34.5 mL of a 25% w/w soln in MeOH, 150.7 mmol, 3.3 equiv) was added dropwise, and the mixture was allowed to reach r.t. over 1 h. This mixture containing intermediate cyanamide adduct 18 was then acidified to ca. pH 2 by the addition of H₂SO₄ and heated at 65 °C for 2.5 h. The mixture was cooled to 0 °C and basified to ca. pH 13 by the dropwise addition of aq NaOH soln (1 M). The resulting slurry was stirred at r.t. for 30 min and filtered. The filter cake was washed with H₂O and dried under vacuum at 60 °C to provide 19 as a white powder (13.63 g, 83%), contaminated with ca. 10% of by-product amine 20 which was removed in the proceeding step. An analytical sample of 19 was isolated by silica gel chromatography (MeOH–CH₂Cl₂, 1:19; gradient).

Mp 223–225 °C.

¹H NMR (400 MHz, DMSO-d ₆): δ = 8.59 (br s, 1 H), 7.37 (d, J = 8.8 Hz, 2 H), 7.26 (d, J = 8.9 Hz, 2 H), 5.56 (br s, 1 H), 3.96–3.89 (m, 5 H), 3.64 (s, 3 H), 3.03 (t, J = 6.8 Hz, 2 H), 1.57 (s, 6 H).

¹³C NMR (100 MHz, DMSO-d ₆): δ = 177.0, 170.9, 165.8, 165.3, 164.9, 142.7, 141.8, 126.5, 126.1, 98.9, 54.9, 52.8, 47.5, 46.5, 31.4, 27.1.

MS (ESI): m/z = 371.2 [M + H]⁺.

HRMS (ESI): m/z [M + H]⁺ calcd for C₁₉H₂₃N₄O₄: 371.1719; found: 371.1722.

2-{4-[4-Amino-2-methoxy-5-oxo-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl]phenyl}-2-methylpropanoic Acid (21)

Ester 19 (15.0 g, 40.5 mmol) was suspended in THF (330 mL) at r.t., and potassium trimethylsilanolate (KOTMS) (17.3 g, 121.5 mmol, 3.0 equiv) was added. The resulting thick suspension was heated at reflux temperature overnight. Next, the reaction volume was reduced by ca. 50% by removing THF to afford a thick, white slurry to which hexane (75 mL) was added. After stirring the slurry for 30 min, the mixture was filtered. The filter cake was washed with H₂O and hexane–EtOAc (1:1, 75 mL) and dried under vacuum to provide acid 21.

Yield: 13.28 g (92%); light-tan powder; mp 240–241 °C.

¹H NMR (400 MHz, DMSO-d ₆): δ = 7.33 (d, J = 8.9 Hz, 2 H), 7.26 (d, J = 8.8 Hz, 2 H), 3.85 (t, J = 6.7 Hz, 2 H), 3.82 (s, 3 H), 2.91 (t, J = 6.7 Hz, 2 H), 1.44 (s, 6 H).

¹³C NMR (100 MHz, DMSO-d ₆): δ = 178.2, 170.9, 165.9, 165.3, 164.9, 143.3, 141.6, 126.6, 126.0, 98.9, 54.9, 47.6, 46.2, 31.4, 27.1.

MS (ESI): m/z = 357.1 [M + H]⁺.

HRMS (ESI): m/z [M + H]⁺ calcd for C₁₈H₂₁N₄O₄: 357.1563; found: 357.1563.

2-{4-[4-Amino-2-methoxy-5-oxo-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl]phenyl}-2-methylpropanamide (2)

Acid 21 (15.0 g, 42.1 mmol) was suspended in DMF (75 mL) and CDI (8.2 g, 50.5 mmol, 1.2 equiv) was added. After stirring for 45 min, 28% aq NH₄OH soln (100 mL) was added in one portion. The resulting thick suspension was stirred at r.t. for 2 h then filtered and washed with H₂O. The crude product was recrystallized from AcOH–H₂O, then dried under vacuum at 60 °C to afford pure target product 2.

Yield: 12.9 g (86%); white powder; mp 245 °C.

¹H NMR (500 MHz, DMSO-d ₆): δ = 8.37 (d, J = 3.9 Hz, 1 H), 7.79 (d, J = 3.9 Hz, 1 H), 7.36 (d, J = 8.8 Hz, 2 H), 7.29 (d, J = 8.8 Hz, 2 H), 6.94 (s, 1 H), 6.90 (s, 1 H), 3.88 (t, J = 6.7 Hz, 2 H), 3.85 (s, 3 H), 2.95 (t, J = 6.8 Hz, 2 H), 1.45 (s, 6 H).

¹³C NMR (125 MHz, DMSO-d ₆): δ = 178.4, 170.9, 165.8, 165.3, 164.9, 144.6, 141.4, 126.7, 125.9, 98.9, 54.9, 47.6, 46.4, 31.4, 27.5.

MS (ESI): m/z = 356.0 [M + H]⁺.

HRMS (ESI): m/z [M + H]⁺ calcd for C₁₈H₂₂N₅O₃: 356.1723; found: 356.1723.

Acknowledgment

We would like to thank Dr. Vincent Mascitti for helpful discussions and suggestions, and Dr. Yun Huang, Mr. David Nelson, and Mr. Dauda Lapido for analytical support.

• References

• 1a Birch AM, Buckett LK, Turnbull AV. Curr. Opin. Drug Discovery Dev. 2010; 13: 489
  Search in Google Scholar
• 1b Zammit VA, Buckett LK, Turnbull AV, Wure H. Pharmacol. Ther. 2008; 118: 295
  CrossrefSearch in Google Scholar
• 2a Dow RL, Li J.-C, Pence MP, Gibbs ME, LaPerle JL, Litchfield J, Piotrowski DW, Munchhof MJ, Manion TB, Zavadoski WJ, Walker GS, McPherson RK, Tapley S, Sugarman E, Guzman-Perez A, DaSilva-Jardine P. ACS Med. Chem. Lett. 2011; 2: 407 ; and references cited therein
  CrossrefSearch in Google Scholar
• 2b Dow RL, Munchhof MJ. US Patent 0197590, 2010
• 2c Aspnes GE, Dow RL, Munchhof MJ. US Patent 0197591, 2010
• 3 To the best of our knowledge, this aminoalkoxypyrimido pyridone scaffold was unprecedented in the chemical literature, prior to the disclosure of such structures in our recent patents. For a related system accessed through N-aryl imines, see: Wada A, Hirai S, Hanaoka M. Chem. Pharm. Bull. 1991; 39: 1189
  CrossrefSearch in Google Scholar
• 4a Surry DS, Buchwald SL. Angew. Chem. Int. Ed. 2008; 47: 6338
  CrossrefPubMedSearch in Google Scholar
• 4b Huang X, Anderson KW, Zim D, Jiang L, Klapars A, Buchwald SL. J. Am. Chem. Soc. 2003; 125: 6653
  CrossrefPubMedSearch in Google Scholar
• 5 X-Phos was chosen as the ligand in this catalyst system based on its commercial availability and low cost, relative to BrettPhos, which specializes in the mono N-arylation of primary amines, see: Fors BP, Watson DA, Biscoe MR, Buchwald SL. J. Am. Chem. Soc. 2008; 130: 13552
  CrossrefSearch in Google Scholar

For similar Dieckmann-type condensations en route to five-membered rings, see:
• 6a Yendapally R, Hurdle JG, Carson EI, Lee RB, Lee RE. J. Med. Chem. 2008; 51: 1487
  CrossrefSearch in Google Scholar
• 6b Kulkarni BA, Ganesan A. Angew. Chem. Int. Ed. 1997; 36: 2454
  Search in Google Scholar
• 7 Several organic and inorganic bases also effected this cyclization, however, the use of DBU consistently resulted in the cleanest and most efficient reactions
• 8 Farr RN, Alabaster RJ, Chung JY. L, Craig R, Edwards JS, Gibson AW, Ho G.-J, Humphrey GR, Johnson SA, Grabowski EJ. J. Tetrahedron: Asymmetry 2003; 14: 3503
  CrossrefSearch in Google Scholar
• 9 This aza-Michael addition proceeded under acidic or basic conditions, both of which afforded aniline 8 with similar conversion profiles
• 10 The use of trimethylsilyldiazomethane proved useful in the mild O-methylation of substrates containing acid-sensitive functionalities incompatible with the oxalyl chloride protocol. Surprisingly, vinylogous carbamates such as 12 were inert to standard O-alkylation conditions (base and methyl iodide or dimethyl sulfate), presumably due to the weak nucleophilicity of the corresponding anions. Attempts to drive Fischer-type esterifications to completion with chemoselectivity and mild conditions were not successful

Based on modifications of procedures reported for de novo syntheses of aminopyrimidines, see:
• 11a Schmidt H.-W, Koitz G, Junek H. J. Heterocycl. Chem. 1987; 24: 1305
  CrossrefSearch in Google Scholar
• 11b Perez MA, Soto JL, Guzman F, Alcala H. J. Chem. Soc., Perkin Trans. 1 1985; 87
  Crossref
• 12 To date, efforts to eliminate completely the formation of amine 20 have not been realized
• 13 For analogue production, the addition of O-alkylisoureas to vinylogous carbamates such as 17 (step f) offered a convenient point of divergence in the rapid exploration of structure–activity relationships around the methoxy substituent of the pyrimidine

• References

• 1a Birch AM, Buckett LK, Turnbull AV. Curr. Opin. Drug Discovery Dev. 2010; 13: 489
  Search in Google Scholar
• 1b Zammit VA, Buckett LK, Turnbull AV, Wure H. Pharmacol. Ther. 2008; 118: 295
  CrossrefSearch in Google Scholar
• 2a Dow RL, Li J.-C, Pence MP, Gibbs ME, LaPerle JL, Litchfield J, Piotrowski DW, Munchhof MJ, Manion TB, Zavadoski WJ, Walker GS, McPherson RK, Tapley S, Sugarman E, Guzman-Perez A, DaSilva-Jardine P. ACS Med. Chem. Lett. 2011; 2: 407 ; and references cited therein
  CrossrefSearch in Google Scholar
• 2b Dow RL, Munchhof MJ. US Patent 0197590, 2010
• 2c Aspnes GE, Dow RL, Munchhof MJ. US Patent 0197591, 2010
• 3 To the best of our knowledge, this aminoalkoxypyrimido pyridone scaffold was unprecedented in the chemical literature, prior to the disclosure of such structures in our recent patents. For a related system accessed through N-aryl imines, see: Wada A, Hirai S, Hanaoka M. Chem. Pharm. Bull. 1991; 39: 1189
  CrossrefSearch in Google Scholar
• 4a Surry DS, Buchwald SL. Angew. Chem. Int. Ed. 2008; 47: 6338
  CrossrefPubMedSearch in Google Scholar
• 4b Huang X, Anderson KW, Zim D, Jiang L, Klapars A, Buchwald SL. J. Am. Chem. Soc. 2003; 125: 6653
  CrossrefPubMedSearch in Google Scholar
• 5 X-Phos was chosen as the ligand in this catalyst system based on its commercial availability and low cost, relative to BrettPhos, which specializes in the mono N-arylation of primary amines, see: Fors BP, Watson DA, Biscoe MR, Buchwald SL. J. Am. Chem. Soc. 2008; 130: 13552
  CrossrefSearch in Google Scholar

For similar Dieckmann-type condensations en route to five-membered rings, see:
• 6a Yendapally R, Hurdle JG, Carson EI, Lee RB, Lee RE. J. Med. Chem. 2008; 51: 1487
  CrossrefSearch in Google Scholar
• 6b Kulkarni BA, Ganesan A. Angew. Chem. Int. Ed. 1997; 36: 2454
  Search in Google Scholar
• 7 Several organic and inorganic bases also effected this cyclization, however, the use of DBU consistently resulted in the cleanest and most efficient reactions
• 8 Farr RN, Alabaster RJ, Chung JY. L, Craig R, Edwards JS, Gibson AW, Ho G.-J, Humphrey GR, Johnson SA, Grabowski EJ. J. Tetrahedron: Asymmetry 2003; 14: 3503
  CrossrefSearch in Google Scholar
• 9 This aza-Michael addition proceeded under acidic or basic conditions, both of which afforded aniline 8 with similar conversion profiles
• 10 The use of trimethylsilyldiazomethane proved useful in the mild O-methylation of substrates containing acid-sensitive functionalities incompatible with the oxalyl chloride protocol. Surprisingly, vinylogous carbamates such as 12 were inert to standard O-alkylation conditions (base and methyl iodide or dimethyl sulfate), presumably due to the weak nucleophilicity of the corresponding anions. Attempts to drive Fischer-type esterifications to completion with chemoselectivity and mild conditions were not successful

Based on modifications of procedures reported for de novo syntheses of aminopyrimidines, see:
• 11a Schmidt H.-W, Koitz G, Junek H. J. Heterocycl. Chem. 1987; 24: 1305
  CrossrefSearch in Google Scholar
• 11b Perez MA, Soto JL, Guzman F, Alcala H. J. Chem. Soc., Perkin Trans. 1 1985; 87
  Crossref
• 12 To date, efforts to eliminate completely the formation of amine 20 have not been realized
• 13 For analogue production, the addition of O-alkylisoureas to vinylogous carbamates such as 17 (step f) offered a convenient point of divergence in the rapid exploration of structure–activity relationships around the methoxy substituent of the pyrimidine