RVX-208 – Acbi1 Chemical

In vitro biosynthesis, isolation, and identiﬁcation of predominant metabolites of 2-(4-(2-hydroxyethoxy)-3,5-dimethylphenyl)-
5,7-dimethoxyquinazolin-4(3H)-one (RVX-208)
Yuri L. Khmelnitsky a, Vadim V. Mozhaev a, Ian C. Cotterill a, Peter C. Michels a, Sihem Boudjabi b, Vladimir Khlebnikov b, M. Madhava Reddy b, Gregory S. Wagner c,
Henrik C. Hansen c,*
a AMRI, 26 Corporate Circle, PO Box 15098, Albany, NY 12212-509, USA
b NAEJA Pharmaceutical Inc., 4290-91A Street, Edmonton, Alberta T6E 5V2, Canada
c Resverlogix Corp. Suite 202, 279 Midpark Way SE, Calgary, Alberta T2X 1M2, Canada

A B S T R A C T

The structures of the two predominant metabolites (M4 and M5) of RVX-208, observed both in in vitro human and animal liver microsomal incubations, as well as in plasma from animal in vivo studies, were determined. A panel of biocatalytic systems was tested to identify biocatalysts suitable for milligram scale production of metabolite M4 from RVX-208. Rabbit liver S9 fraction was selected as the most suitable system, primarily based on pragmatic metrics such as catalyst cost and estimated yield of M4 (w55%). Glucuronidation of RVX-208 catalyzed by rabbit liver S9 fraction was optimized to produce M4 in amounts sufﬁcient for structural characterization. Structural studies using LC/MS/MS analysis and 1H NMR spectroscopy showed the formation of a glycosidic bond between the primary hydroxyl group of RVX-208 and glucuronic acid. NMR results suggested that the glycosidic bond has the b-anomeric conﬁguration. A synthetic sample of M4 conﬁrmed the proposed structure. Metabolite M5, hypothesized to be the carboxylate of RVX-208, was prepared using human liver microsomes, puriﬁed by HPLC, and characterized by LC/MS/MS and 1H NMR. The structure was conﬁrmed by comparison to a synthetic sample. Both samples conﬁrmed M5 as a product of oxidation of primary hydroxyl group of RVX-208 to carboxylic acid.

1. Introduction

2-(4-(2-Hydroxyethoxy)-3,5-dimethylphenyl)-5,7- dimethoxyquinazolin-4(3H)-one (RVX-208) (Fig. 1) is a novel BET bromodomain inhibitor. It is an orally active small molecule drug candidate for the treatment of cardiovascular diseases. It is pro- posed to work by increasing the levels of apolipoprotein A1 (Apo- A1), resulting in an increase in high-density lipoprotein cholesterol (HDL-c), thereby potentially reducing the risk for cardiovascular disease [1,2]. RVX-208 is currently in phase 2 clinical trials for the treatment of atherosclerosis. Identiﬁcation of the main metabolites is an essential component of understanding the elimination of the parent compound from the body, as well as understanding the safety proﬁle for the drug candidate.

* Corresponding author. Tel.: þ1 403 254 9252; fax: þ1 403 256 8495.
E-mail address: [email protected] (H.C. Hansen).

In preliminary in vitro metabolite proﬁling and identiﬁcation studies using liver fractions from various mammalian species, it was determined that the metabolic pathways of RVX-208 involved the formation of two signiﬁcant metabolites, a glucuronidated metabolite M4 and an oxidized metabolite M5. Metabolites M4 and M5 were also identiﬁed as predominant metabolites in follow-up in vivo animal studies (unpublished results). Therefore, it was important to establish the exact structures of these me- tabolites, which required synthesis of authentic M4 and M5 suit- able for NMR structural characterization. In the absence of prior knowledge of metabolite structures, one of the most effective strategies for the production of metabolites is the use of the actual liver metabolic enzyme systems, typically found in hepatocytes, microsomes, or S9 fractions [3e5]. We applied this biosynthetic strategy to synthesize and characterize the two predominant metabolites (M4 and M5) of RVX-208 to support the ongoing development program.

Table 2
Estimated yields (%) of metabolite M4 in incubations of RVX-208 with S9 Fractions.a

<1%). Based on results of the screen, microsomes and S9 fraction from rabbit liver and microsomes from calf liver all had an esti- mated yield above 30% and were selected for further optimization. Fig. 1. Structures of RVX-208 and its metabolites M4 and M5. 2. Results and discussion 2.1. Synthesis and structural characterization of glucuronidated metabolite M4 2.1.1. Screening of biocatalytic systems for production of metabolite M4 Initially, a broad spectrum of mammalian liver microsomes (human, mouse, rat, dog, rabbit, monkey, minipig, calf), S9 fractions (from human, mouse, rat, dog, rabbit, monkey, minipig), and twelve human recombinant glucuronyl transferases were screened to identify a system that would be practical for the biosynthesis of M4. The initial selection criteria was the estimated yield after 24 h of incubation. Estimated yields of M4 in reactions catalyzed by mammalian liver microsomes and S9 fractions are shown in Tables 1 and 2, respectively. Metabolite yields were estimated from integrated areas of corresponding peaks in HPLC chromatograms with UV detection at 230 nm. In reactions with UGT enzymes, metabolite M4 was formed only in trace amounts (estimated yield Table 1 Estimated yields (%) of metabolite M4 in incubations of RVX-208 with liver microsomes.a Source of microsomes Reaction time (hours) 1 6 24 Male Beagle Dog 0.9 6.4 15.5 Male Cynomolgus Monkey 0.2 1.1 2.9 male Sprague-dawley rat 0.1 0.5 0.7 2.1.2. Synthesis, puriﬁcation and structural characterization of metabolite M4 The yield in the glucuronide metabolite synthesis primarily depends on the concentrations of the parent compound and UDPGA (the cofactor for glucuronidation), and the amount of a catalyst added. In the optimization screen, these parameters were varied over a broad range of values at a ﬁxed reaction time of 24 h. The results are shown in Table 3. In the majority of reaction conditions tested in the optimization screen, the estimated yield of M4 exceeded 20%, and in several reaction systems it was greater than 50%. Overall, calf microsomes were better catalysts than rabbit microsomes and S9 fraction. However, because rabbit liver S9 fraction offers the most cost effective option compared to other catalysts in Table 3, they were selected as a practical catalyst for synthesis of M4. Based on the results of the optimization study, the following reaction conditions were selected for scale-up synthesis of metabolite M4:1 mM RVX- 208, 2 mg/mL S9 protein (rabbit), and 10 mM UDPGA. After 24 h of incubation LC/MS analysis of the reaction mixture showed a 22% yield of M4. Following work-up of the reaction and puriﬁcation by preparative HPLC, a total of 2.6 mg of M4 was ob- tained as a white solid with >99% purity as determined by HPLC (see Supporting information).

Table 3
Optimization of reaction conditions for synthesis of metabolite M4.

Catalyst [UDPGA] [Total protein] [RVX-208] Estimated mM mg/mL mM yield, %

Rabbit microsomes 5 4 2 30.0
0.5 24.1
1 2 38.4
0.5 39.5
0.2 34.6
25 4 2 48.9
0.5 54.3
1 2 39.7
0.5 49.4
0.2 49.7
Rabbit S9 fraction 5 4 2 30.3
0.5 33.0
1 2 17.5
0.5 23.2
0.2 22.3
25 4 2 35.2
0.5 41.3

Male ICR/CD-1 Mouse 0.1 0.3 0.5 1 2 16.7
Male New Zealand White Rabbit 12.2 22.4 30.3 0.5 25.7
Pooled Human Mixed Sexes (10 individuals) 0.1 0.6 1.2 0.2 21.4
Calf 3.7 27.7 35.9 Calf microsomes 5 4 2 50.1
Male Yucatan Minipig 2.5 11.8 14.3 1 2 40.9
a Estimated from the HPLC peak area. 0.5 56.0

Structural characterization of metabolite M4 was conducted using LC/MS/MS analysis and 1H NMR spectroscopy. Results of LC/ MS/MS analysis suggested that glucuronidation occurred at the hydroxyl group of RVX-208 (Fig. 2). 1H NMR data analysis conﬁrmed the formation of a glycosidic bond between the hydroxyl group of RVX-208 and glucuronic acid (See Supporting information). Analysis of the 1H NMR spectrum in narrow spec- tral ranges (Fig. 3) allowed determination of the anomeric conﬁg- uration of the glycosidic bond. Based on the coupling constant for the anomeric proton, which is approximately 8 Hz (panel B in Fig. 3), the glycosidic bond was determined to be in the b-conﬁg- uration [6]. The proposed structure of metabolite M4, derived from LC/MS/MS and NMR data, is shown in Fig. 1.
To verify the structure of metabolite M4, it was synthesized chemically from RVX-208 using the synthetic route shown in Scheme 1 (see Experimental section for details). For the synthesis, the initial protection of the amide nitrogen was required to avoid formation of the N-glucuronide. Having the 2-O-acetate glycosyl donor supported the formation of the b-anomer. Both chemically and biosynthetically prepared samples of M4 exhibit identical LC/ MS/MS and 1H NMR data, conﬁrming the structure shown in Fig. 1.

2.2. Synthesis and structural characterization of oxidized metabolite M5

For structural identiﬁcation of metabolite M5, conditions of the initial screening protocol used for microsomal incubations were found to be adequate to generate sufﬁcient quantities of M5. Using human liver microsomes an analytical amount of M5 was produced from RVX-208 (See Experimental section for details). After puriﬁ- cation by HPLC, the material was subjected to LC/MS/MS charac- terization. Results of this analysis (Fig. 4) suggested that M5 is the result of oxidation of the primary alcohol in RVX-208 to a carboxylic acid. To verify the structure of metabolite M5, it was synthesized

chemically using the synthetic route shown in Scheme 2 (See Experimental for details). The LC/MS/MS and 1H NMR properties of synthetic M5 are identical to those of the metabolite M5 prepared using human liver microsomes, thus conﬁrming the structural identity of M5.

3. Conclusion

In conclusion, we successfully applied biosynthetic strategy to synthesize the predominant metabolites (M4 and M5) of the clin- ical drug candidate RVX-208 on a preparative scale. A wide variety of biocatalytic systems were tested for synthetic suitability, the most effective catalysts were identiﬁed, and conditions were opti- mized to achieve practical yields of the target metabolites. The structures of the metabolites were proposed from LC/MS/MS and 1H NMR spectroscopy data, and they were veriﬁed by direct chemical synthesis.

4. Experimental procedures

4.1. General procedures and materials

Calf liver microsomes were obtained from CellzDirect (Pittsboro, NC). All other mammalian liver microsomes and mammalian S9 fractions were obtained from In Vitro Technologies, Inc. (Baltimore, MD). Metabolic competency of liver microsomes and S9 fractions was conﬁrmed in a separate control experiment using a standard substrate, testosterone. Components of the NADPH regeneration system and human recombinant UDP-glucuronyl transferases were obtained from BD Gentest (Woburn, MA). RVX-208 was supplied by Resverlogix Corp. (Calgary, Canada). Solvents and reagents for chemical syntheses were used as purchased from commercial suppliers, unless otherwise noted. Synthetic chemical reactions were monitored by thin-layer chromatography (TLC) carried out on

Fig. 2. MS/MS spectrum and fragmentation analysis for metabolite M4.

Fig. 3. 1H NMR spectra of puriﬁed metabolite M4, showing narrow spectral ranges. Panels: A, spectrum in the range from 6.0 to 8.0 ppm; B, spectrum in the range from 3.8 to
4.5 ppm; C, spectrum in the range from 3.35 to 3.8 ppm.

MeO

N
NH
MeO O
1

O OH

MeO

MeO O O
2

OAc
AcO OAc
OH
Br O CO2Me
3

RVX-208

AcO OAc

HO OH

MeO

MeO O O

O CO2Me

MeO

MeO O

O CO2H

4 M4

Scheme 1. Chemical synthesis of metabolite M4 (see Materials and methods for details). a: NaH, DMF, chloromethylpivalate; b: methylene chloride, AgO; c: NaOMe, MeOH.

MachereyeNagel (MN) Alugram SIL G/UV254 silica gel 60 plates with ﬂuorescent indicator UV254 (catalog #818 133) or Sigmae Aldrich silica gel plates (catalog #Z193291-1PAK) using phospho- molybdic acid in absolute ethanol as a developing agent. NMR spectra were recorded on a 400 MHz Varian Mercury spectrometer or a 500 MHz Bruker Avance System, using residual undeuterated solvent as an internal reference. The low resolution mass spectra (MS) were recorded on a Water Micromass ZQ using electrospray ionization-ion trap (ESI) unless otherwise stated. MS/MS analysis was carried out on an API 3200 system (Applied Biosystems). Q1 was set up to select a precursor ion at desired m/z, pass the ion through a hexapole collision cell, and the Q3 analyzer then scanned the product ions in the mass range from 0 to 600 amu. Nitrogen was used as the collision gas, and the collision energy was scanned to 100 V.

4.1.1. Reactions with mammalian liver microsomes and S9 fractions
A 10 mM stock solution of RVX-208 [7] in DMSO was prepared and used in these experiments. The incubation mixtures, prepared in duplicate, consisted of 0.50 mg microsomal protein, 100 mM potassium phosphate buffer, pH 7.4, 100 mM test compound, and the NADPH regenerating system (1.3 mM NADPþ, 3.3 mM glucose- 6-phosphate, 0.4 U/mL glucose-6-phosphate dehydrogenase, and
3.3 mM magnesium chloride), UDPGA (5 mM), and D-saccharic acid lactone (5 mM) in a ﬁnal volume of 0.5 mL. The reaction mixture
minus the microsomal protein or S9 fraction was pre-warmed for 5 min at 37 ◦C. In experiments with S9 fractions, 0.1 mL of the S9 fraction suspension was added directly to the reaction mixture. In
experiments with microsomes, prior to addition to the reaction mixture, ice-cold microsomes were pre-incubated for 20 min with alamethicin (4 mL of alamethicin stock solution in DMSO, 1.5 mg/

Fig. 4. LC/MS/MS and fragmentation analysis of puriﬁed metabolite M5. Panels: A, UV trace (detection at 230 nm); B, MS/MS spectrum.

OHC

a OHC
5 6

to 100% B in 20 min, followed by an isocratic hold for 1 min at a
ﬂow rate of 0.35 mL/min. The column was re-equilibrated for
OH 10 min after programming back to the starting solvent mixture over 0.5 min. UV data were acquired at 230 nm with an injection
volume of 10 mL of sample solution. The mass spectrometer was operated in positive ion mode with spray voltage set at 5300 V. The ion source temperature was set at 425 ◦C. The 0.35 mL/min
OH efﬂuent from the HPLC column was directed to an ESI ion source

MeO 7 O
b

2 MeO

MeO O M5

without splitting after UV detection. The mass scan was performed in the range from 200 to 750 amu. The retention time of RVX-208 was 12.3 min.

4.1.4. LC/MS analysis for monitoring testosterone-based reactions

Scheme 2. Chemical synthesis of metabolite M5 (see Materials and methods for de- tails). a: Bromoacetic acid, NaOH, water; b: NaHSO3, pTSA, DMAC.
30 mL, were mixed with 0.2 mL of microsomal suspension) and then added to initiate the reaction. After 1 h, 6 h, or 24 h incubation in a shaking water bath at 37 ◦C, the reactions were stopped by the
addition of an equal volume of methanol. After incubation on ice for 15 min, samples were centrifuged (10,000g, 15 min, 4 ◦C) to remove precipitated protein. The supernatants were ﬁltered using a Teﬂon® syringe ﬁlter (4 mm PTFE, 0.45 mm) from Whatman, Inc. (Clifton, NJ) followed by evaporation to dryness in a Centrivap concentrator and Freeze Dry system (both Labconco, Kansas City, MO). Sample resi- dues were resuspended in 80 mL of 50% (v/v) methanol, mixed for
15 s on a vortex mixer, and sonicated for 30 s followed by centri- fugation (20,000g, 30 s, room temperature). The supernatant was then transferred to HPLC vials for LC/MS analysis.
In experiments to optimize the reaction conditions for the synthesis of metabolite M4, stock solutions of RVX-208 in DMSO with concentrations of 0.2 M, 0.05 M, and 0.02 M were prepared and used to obtain ﬁnal substrate concentrations of 2 mM, 0.5 mM, and 0.2 mM, respectively, after 100-fold dilution. The concentration of microsomal protein in the reaction mixture was 1 mg/mL or 4 mg/mL; the concentration of UDPGA was 5 mM or 25 mM.

4.1.2. Reactions with recombinant UDP-glucuronyl transferases
A 10 mM stock solution of RVX-208 in methanol was prepared and used in these experiments. Incubation mixtures consisted of 100 mM potassium phosphate buffer, pH 7.4, 100 mM test com- pound, 3.3 mM magnesium chloride, 5 mM UDPGA, and protein (from 0.025 mg/mL to 1 mg/mL, depending on the enzyme) in a ﬁnal volume of 0.5 mL. The reaction mixture minus the enzyme was
pre-warmed for 5 min at 37 ◦C and the enzyme suspension was
added. After 1 h, 6 h, or 24 h incubation in a shaking water bath at 37 ◦C, the reactions were stopped by the addition of an equal vol-
ume of methanol. After incubation on ice for 15 min, samples were centrifuged (10,000g, 15 min, 4 ◦C) to remove precipitated protein. The supernatants were ﬁltered using a Teﬂon® syringe ﬁlter (4 mm PTFE, 0.45 mm) from Whatman, Inc. (Clifton, NJ) followed by evaporation to dryness in a Centrivap concentrator and Freeze Dry system (both Labconco, Kansas City, MO). Sample residues were resuspended in 80 mL of 50% (v/v) methanol, mixed for 15 s on a
vortex mixer, and sonicated for 30 s followed by centrifugation (20,000g, 30 s, room temperature). The supernatant was then transferred to HPLC vials for LC/MS analysis.

4.1.3. LC/MS analysis for monitoring RVX-208 and its metabolites
LC/MS analysis was performed on a PE SCIEX API 150 system with a PDA detector. Chromatography was done at room temper- ature using a Waters SunFire C18 column (100 2.1 mm, 3.5 mm) with the mobile phase initially composed of 98% of solvent A (0.1% formic acid in water) and 2% of solvent B (0.1% formic acid in acetonitrile). Elution was performed using a linear gradient from 2

LC/MS analysis was performed on a PE SCIEX API 2000 system with a PDA detector. Chromatography was done at room temper- ature using a Waters SunFire C18 column (2.1 50 mm, 3.5 mm) with the mobile phase initially composed of 10% of solvent A (0.1% formic acid in water) and 90% of solvent B (0.1% formic acid in acetonitrile). Elution was performed with a linear gradient from 10 to 30% B in 12 min, then from 30 to 100% B in 0.5 min, and isocratic hold for 2 min at a ﬂow rate of 0.35 mL/min. The column was re- equilibrated for 8 min after programming back to the starting sol- vent mixture over 0.5 min. UV data were acquired with total wavelength count (TWC) and the injection volume was 20 mL of sample solution. The mass spectrometer was operated in positive
ion mode with spray voltage set at 5400 V. The ion source tem- perature was set at 450 ◦C. The 350 mL/min efﬂuent from the HPLC column was directed to an ESI ion source without splitting after UV detection. The retention time of 6b-hydroxytestosterone was
13.8 min.

4.1.5. Biocatalytic synthesis and puriﬁcation of metabolite M4
All experiments were carried out with the 0.1 M stock solution of RVX-208 in DMSO. Six identical incubations were performed in parallel in 20 mL scintillation vials, each containing 10 mL of the reaction mixture. Each reaction consisted of 20 mg rabbit liver S9 protein, 100 mM potassium phosphate buffer, pH 7.4, 1 mM test compound, the NADPH regenerating system (1.3 mM NADPþ,
3.3 mM glucose-6-phosphate, 0.4 U/mL glucose-6-phosphate de-
hydrogenase, and 3.3 mM magnesium chloride), and 10 mM UDPGA in a ﬁnal volume of 10 mL. After 24 h incubation in a shaker at 37 ◦C, the reaction mixtures were combined, diluted 2-fold with
cold methanol, centrifuged, and the supernatant was diluted ﬁve- fold with water and passed over a solvated and equilibrated All- tech C18 SPE (4 g) cartridge. The column was washed with 20 mL of acetonitrile and then equilibrated with 20 mL of water. Following sample loading, the absorbent was washed with 40 mL of water. The elution was then performed using acetonitrile (3 20 mL), which allowed complete desorption of RVX-208, whereas metab- olite M4 was not eluted. Final elution was performed using 8 mL of 50% (v/v) acetonitrile. The eluate was then injected in 4 mL portions on a Shimadzu LC8A preparative HPLC system consisting of two Shimadzu LC8A pumps, a Shimadzu SPD-10A UV detector, and a SCL-10A system controller. Chromatography was accomplished at room temperature using a Waters Sunﬁre Prep C18 OBD column (150 19 mm, 5 mm) with the mobile phase initially composed of 98% of solvent A (water with 0.1% TFA) and 2% of solvent B (acetonitrile with 0.1% TFA). Elution was performed with a linear gradient from 2 to 100% B in 20 min followed by an isocratic hold at 100% B for 1 min at a ﬂow rate of 20 mL/min. The column was re- equilibrated for 10 min after programming back to the starting solvent mixture over 0.5 min. UV data were acquired at 230 nm. Fractions were collected for the peak eluting at 9.5 min. The pooled sample was concentrated to dryness ﬁrst by evaporating acetoni- trile under reduced pressure and then removing the residual water

by lyophilization on a Labconco freeze dryer to give 2.6 mg of M4 as a white solid.

4.1.6. Biocatalytic synthesis and puriﬁcation of metabolite M5
The reaction of RVX-208 with human liver microsomes was set up as described above in the screening protocol. After 4 h incuba- tion at 37 ◦C, the reaction mixture was diluted 2-fold with cold
methanol, incubated on ice for 15 min, and centrifuged (10,000g, 15 min, 4 ◦C) to remove precipitated protein. The supernatant was ﬁltered using a Teﬂon® syringe ﬁlter (4 mm PTFE, 0.45 mm) from Whatman, Inc. (Clifton, NJ) and injected (20 mL) on an analytical HPLC using the same gradient as described above for LC/MS anal-
ysis of RVX-208 and its metabolites. The fractions corresponding to metabolite M5 (retention time 13.5 min) were collected and sub- mitted to MS/MS analysis.

4.2. Chemical synthesis

4.2.1. 2,2-Dimethylpropionic acid 2-(4-(2-hydroxyethoxy)-3,5- dimethylphenyl)-5,7-dimethoxy-4-oxo-4H-quinazolin-3-ylmethyl ester (2)
To a stirred solution of 2-(4-(2-hydroxyethoxy)-3,5-dimeth ylphenyl)-5,7-dimethoxy-3H-quinazolin-4-one (1; RVX-208) (3.0 g
, 8.1 mmol) in DMF (75 mL) under nitrogen atmosphere, sodium hydride (60% in mineral oil, 0.39 g, 9.7 mmol) was added in small portions at room temperature. The reaction mixture was stirred for 1 h, then chloromethylpivalate (1.76 mL, 12.1 mmol) was added, and stirring was continued at room temperature for 24 h. The reaction was quenched with water and the mixture was extracted with ethyl acetate. The crude material was puriﬁed by column chromatography (Silica Gel 230e400 mesh; 7/3 methylene chloride/ethyl acetate as eluent; column eluted until compound was isolated as monitored by TLC) to give 2,2-dimethylpropionic acid 2-(4-(2-hydroxyethoxy)- 3,5-dimethylphenyl)-5,7-dimethoxy-4-oxo-4H-quinazolin-3- ylmethyl ester (2) as a white solid. Yield: 2.2 g (56%). 1H NMR (400 MHz, CDCl3): d 8.20 (s, 2H), 6.95 (d, J ¼ 2.4 Hz, 1H), 6.47 (d,
J ¼ 2.4 Hz, 1H), 6.39 (s, 2H), 3.97 (m, 4H), 3.95 (s, 3H), 3.93 (s, 3H),
2.40 (s, 6H), 2.22 (t, J ¼ 5.8 Hz, 1H), 1.21 (s, 9H).
4.2.2. 3,4,5-Triacetoxy-6-(2-(4-(3-(2,2-
dimethylpropionyloxymethyl)-5,7-dimethoxy-4-oxo-3,4- dihydroquinazolin-2-yl)-2,6-dimethylphenoxy)ethoxy)- tetrahydropyran-2-carboxylic acid methyl ester (4)
To a stirred solution of 2,2-dimethylpropionic acid 2-(4-(2- hydroxyethoxy)-3,5-dimethylphenyl)-5,7-dimethoxy-4-oxo-4H-qui
nazolin-3-ylmethyl ester (2) (3.10 g, 6.40 mmol) in methylene chlo-
ride (76 mL), 4 ˚A molecular sieves (15.4 g) were added under nitro- gen. The resulting mixture was stirred for 5 min before commercially
available (2S,3S,4S,5R,6R)-3,4,5-triacetoxy-6-bromotetrahydropyra n-2-carboxylic acid methyl ester (3) (2.54 g, 6.40 mmol) was added followed by silver (I) oxide (4.50 g, 19.3 mmol). The reaction mixture was stirred at room temperature in the dark for 3 days. The solid was ﬁltered off and was washed with methylene chloride. The combined ﬁltrates were concentrated and the crude material was puriﬁed by column chromatography (Silica Gel 230e400 mesh; 7/3 methylene chloride/ethyl acetate as eluent; column eluted until compound was isolated as monitored by TLC) followed by crystallization from acetone/hexanes (15 mL/25 mL) to give 3,4,5-triacetoxy-6-(2-(4-(3- (2,2-dimethylpropionyloxymethyl)-5,7-dimethoxy-4-oxo-3,4-dihy droquinazolin-2-yl)-2,6-dimethylphenoxy)ethoxy)tetrahydropyran
-2-carboxylic acid methyl ester (4) as a white solid. Yield: 1.7 g (34%).
1H NMR (400 MHz, CDCl3): d 8.18 (s, 2H), 6.95 (d, J ¼ 2.2 Hz, 1H), 6.47
(d, J ¼ 2.2 Hz, 1H), 6.39 (s, 2H), 5.33e5.24 (m, 2H), 5.14e5.07 (m, 1H),
4.79 (d, J ¼ 7.6 Hz, 1H), 4.21e4.12 (m, 1H), 4.04e3.96 (m, 3H), 3.95 (s,

3H), 3.93 (s, 3H), 3.88e3.85 (m, 1H), 3.77 (s, 3H), 2.36 (s, 6H), 2.05 (s,
3H), 2.03 (s, 6H), 1.21 (s, 9H).

4.2.3. 6-(2-(4-(5,7-Dimethoxy-4-oxo-3,4-dihydroquinazolin-2-yl)- 2,6-dimethylphenoxy)-ethoxy)-3,4,5-trihydroxytetrahydropyran-2- carboxylic acid (M4)
A 0.05 M solution of sodium methoxide in methanol (129 mL,
6.40 mmol) was added to 3,4,5-triacetoxy-6-(2-(4-(3-(2,2- dimethylpropionyloxymethyl)-5,7-dimethoxy-4-oxo-3,4-dihydro quinazolin-2-yl)-2,6-dimethylphenoxy)ethoxy)tetrahydropyran-2-
carboxylic acid methyl ester (4) (0.90 g, 1.1 mmol) at 0 ◦C under
nitrogen. The reaction progress was monitored by TLC and mass spectroscopy. After disappearance of the starting material, water (0.5 mL) was added to hydrolyze the methyl ester over 2e3 h at room temperature. The reaction mixture was treated with amberlite Hþ until pH became neutral. The amberlite was ﬁltered off and
washed with methanol. The crude material was puriﬁed by Biotage
reverse phase column chromatography (KP-C18-HS, 35e70 mm, 90 A; 25/75 to 40/60 gradient of MeOH/H2O as an eluent; column eluted until compound was isolated as monitored by TLC) to give 6- (2-(4-(5,7-dimethoxy-4-oxo-3,4-dihydroquinazolin-2-yl)-2,6-dim ethylphenoxy)ethoxy)-3,4,5-trihydroxytetrahydropyran-2-carbox
ylic acid (M4) as a white solid. Yield: 186 mg (34%). Mp 208.6e 210.8 ◦C; UV (l, log ε): 211 (1.70), 258 (1.68) nm; Purity by HPLC:
97.7%; 1H NMR (400 MHz, CD3OD): d 7.70 (s, 2H), 6.79 (d, J 2.2 Hz,
1H), 6.55 (d, J 2.2 Hz, 1H), 4.42 (d, J 7.7 Hz, 1H), 4.32e4.25 (m, 1H),
4.16e4.09 (m, 1H), 4.08e4.03 (m, 1H), 4.02e3.95 (m, 1H), 3.92 (s,
3H), 3.91 (s, 3H), 3.59 (d, J ¼ 9.0 Hz, 1H), 3.50e3.40 (m, 1H), 3.28 (d,
J ¼ 9.0 Hz, 1H), 2.40 (s, 6H). Exchangeable protons (CO2H, OH and NH) are not detected. MS (ESþ) m/z: 547.24 (Mþ1). Analysis calcu- lated for C26H30N2O11$4H2O (618.52), %: C 50.48; H 6.19; N 4.53.
Found, %: C 50.11; H 5.55, N 4.80. The material (M4) was hydroscopic and prolonged drying did not reduce water content. Water content by Karl Fisher varied between 5.5 and 10.7%.

4.2.4. 4-Formyl-2,6-dimethylphenoxyacetic acid (6)
A solution of sodium hydroxide (2.5 g, 63 mmol) in water (65 mL) was added to a mixture of bromoacetic acid (5.3 g,
38 mmol) and 3,5-dimethyl-4-hydroxybenzaldehyde (5) (1.9 g, 13 mmol) in water (30 mL). The reaction mixture was stirred at 100 ◦C for 24 h, then the solution was acidiﬁed (pH w2) with conc.
HCl. The resulting brown solid was isolated, washed with water, dried, and puriﬁed by column chromatography (Silica Gel 230e 400 mesh; 0e10% gradient of methanol in methylene chloride as an eluent; column eluted until compound was isolated as monitored by TLC) to give 4-formyl-2,6-dimethylphenoxyacetic acid (6) as a light brown solid. Yield 0.40 g (15%). 1H NMR (400 MHz, CDCl3): d 9.90 (s, 1H), 7.59 (s, 2H), 4.54 (s, 2H), 2.39 (s, 6H).

4.2.5. 2-(4-(5,7-Dimethoxy-4-oxo-3,4-dihydroquinazolin-2-yl)- 2,6-dimethylphenoxy)-acetic acid (M5)
To a solution of 2-amino-4,6-dimethoxybenzamide (7) (0.15 g,
0.76 mmol) in DMAC (5 mL) were added 4-formyl-2,6-dimethyl- phenoxyacetic acid (6) (0.16 g, 0.76 mmol), sodium hydrogen sulﬁte (Assay > 58.5%, 0.150 g, 0.84 mmol) and p-toluenesulfonic acid monohydrate (15 mg, 0.076 mmol). The reaction mixture was
stirred at 150 ◦C for 3 h, cooled to room temperature, and water (40 mL) was added. The yellow precipitate was ﬁltered, washed
with water and small amount of methanol, then triturated with 10% methanol in diethyl ether to give 0.084 g of compound which was further puriﬁed by preparative HPLC to give 47 mg (16%) of M5 as a
white solid. Selected data for M5: Mp 281e282 ◦C; UV (l, log ε): 209
(2.58), 261 (2.80) nm; Purity by HPLC: 96.2%; 1H NMR (400 MHz, DMSO-d6): d 12.96 (br s, 1H), 11.85 (s, 1H), 7.90 (s, 2H), 6.74 (d,
J ¼ 2.2 Hz, 1H), 6.52 (d, J ¼ 2.2 Hz, 1H), 4.46 (s, 2H), 3.89 (s, 3H), 3.84

(s, 3H), 2.31 (s, 6H). 13C NMR (100 MHz, DMSO-d6): d 170.8, 164.9,
161.7, 160.4, 158.6, 153.8, 153.1, 131.3 (2 carbons), 129.0 (2 carbons),
128.3, 105.4, 101.9, 98.3, 69.5, 56.6, 56.3, 16.8 (2 carbons).
Acknowledgments

This research was sponsored by Resverlogix Corp. We thank Dr.
Charles Jensen for critically reviewing the manuscript.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.ejmech.2013.03.062.

References

[1] D. Bailey, R. Jahagirdar, A. Gordon, A. Haﬁane, S. Campbell, S. Chatur,
G.S. Wagner, H.C. Hansen, F.S. Chiacchia, J. Johansson, L. Krimbou, N.C.W. Wong,
J. Genest, J. Am. Coll. Cardiol. 55 (2010) 2580e2589.
[2] S.J. Nicholls, A. Gordon, J. Johansson, K. Wolski, C.M. Ballantyne,
J.J.P. Kastelein, A. Taylor, M. Borgman, S.E. Nissen, J. Am. Coll. Cardiol. 57 (2011) 1111e1119.
[3] V.V. Mozhaev, L.V. Mozhaeva, P.C. Michels, Y.L. Khmelnitsky, Drug Metab. Dis- pos. 36 (2008) 1998e2004.
[4] J.F. Levesque, E. Templeton, L. Trimble, C. Berthelette, N. Chauret, Anal. Chem. 77 (2005) 3164e3172.
[5] Q. Wang, R. Jia, C. Ye, M. Garcia, J.B. Li, I.J. Hidalgo, In Vitro Cell. Dev. Biol. Anim. 41 (2005) 97e103.
[6] M.D. Green, T.R. Tephly, Drug Metab. Dispos. 24 (1996) 356e363.
[7] H.C. Hansen, US Patent 8,053,440.