Introduction

Acyl-CoA:cholesterol acyltransferase (ACAT, EC 2.3.1.26) is a promising therapeutic target for the treatment or prevention of hypercholesterolemia and atherosclerosis.1, 2, 3 Although a large number of synthetic ACAT inhibitors were developed in 1980s and 1990s, they caused numerous side effects including diarrhea and adrenal gland toxicity.4, 5, 6 Interestingly, two ACAT isozymes (ACAT1 and ACAT2) with distinct patterns of expression in tissues and physiological functions were discovered in the late 1990s.7, 8, 9, 10, 11 In parallel with the findings, non-selective ACAT inhibitors, avasimibe and pactimibe, which inhibit both ACAT1 and ACA2, were developed. Unfortunately, they did not appear to limit the progression of atherosclerosis in clinical trials.12, 13, 14

Pyripyropene A (PPPA, 1, Figure 1), isolated from the culture broth of Aspergillus fumigatus FO-1289,15, 16, 17 was found to be the first inhibitor highly selective toward ACAT2 isozyme.18, 19, 20 Among approximately 200 PPPA derivatives we had prepared in 1990s (the first generation derivatives),21, 22, 23 1 showed the highest selectivity toward ACAT2.24 Accordingly, 1 was chosen as the best candidate for investigating whether ACAT2 selective inhibition exerts in vivo atheroprotective activity. The in vivo experiments appeared worthy because no in vivo efficacy had been tested using ACAT2 selective inhibitors. As a result, we demonstrated that 1 was orally active in in vivo mouse studies, reducing total cholesterol levels in plasma and atherosclerotic lesion area in mouse models without causing side effects.25

Figure 1
figure 1

The structure of pyripyropene A.

Full size image

As a next step for further development, synthesis of new PPPA derivatives with more potent and more selective activity against ACAT2 than 1 was required. For this purpose, potential in vitro metabolism of 1 in plasmas and liver microsomes from various animal samples including mice and humans was investigated.

In this study, PPPA metabolites were identified by ultra fast liquid chromatography (UFLC) and liquid chromatography/tandem mass spectrometry (LC-MS) in comparison with the authentic derivatives chemically semi-synthesized. Finally, potential metabolic pathways of 1 in the liver and plasma were deduced. Furthermore, ACAT inhibitory activity of the metabolites was evaluated in our established cell-based assay.

Results

Synthesis and UFLC analysis of PPPA derivatives

Esters, amides and thioester residues could be hydrolyzed by carboxyesterase, amide hydrolase and thioesterase, respectively, which show ubiquitous expression with high levels in various tissues,26 so that potential metabolites of 1 can be deduced. Therefore, we synthesized possible PPPA derivatives; 7-deacetyl PPPA (3), 1,7,11-trideacetyl PPPA (8) and N-oxidized PPPA (10). Furthermore, as shown in Scheme 1, other novel PPPA derivatives 2, 4, 6 and 9 were synthesized. 1,7,11-Trideacetyl PPPA (8) was treated with (tBu)2Si(OTf)2 followed by acetylation at 7-hydroxy group to afford 12. Regioselective mono-deprotection of 12 in the presence of NH4F afforded 13 as the mono-deprotected 14 in 78% yield along with small amounts of regioisomer 14 (5.0%) and 1,11-dideacetyl PPPA (6) (5.0%). TEMPO (2,2,6,6-tetramethylpiperidine 1-oxyl)-mediated regioselective oxidation of 13 followed by desilylation at 1-position gave 9 in 47% yield from 13. An acetylation/deprotection sequences of 13 and 14 afforded 1-deacetyl PPPA (2) and 11-deacetyl PPPA (4) in 90% and 84% yield (two steps), respectively. Furthermore, synthesis of the dideacetyl PPPA derivatives 5 and 7 was carried out as illustrated in Scheme 2. Protection of the 7-hydroxyl group of 11 by allyloxycarbonyl (alloc) group followed by the regioselective mono-deprotection of silylene acetal 16 gave as a major product along with a small account of regioisomer 17. Finally, acetylation and deprotection of silylether and alloc group of 16 and 17 afforded 1,7-dideacetyl PPPA (5) and 7,11-dideacetyl PPPA (7) in three steps, respectively. Details of spectral data of PPPA derivatives are available from the Supporting Data.

Next, synthetic PPPA derivatives were analyzed by UFLC and LC-MS with an octadecylsilane column. Under the routine conditions, 1 to 10 were eluted at the respective retention times of 3.8, 3.1, 3.0, 3.5, 2.2, 2.8, 2.6, 1.9, 2.9 and 3.6 min by UFLC and of 11.5 ([M+H]+ 584), 8.6 (542), 7.4 (542), 10.6 (542), 3.3 (500), 5.9 (500), 4.7 (500), 2.7 (458), 7.3 (498) and 10.5 min (600) by LC-MS (Table 1).

Table 1 Retention times and mass data of authentic PPPA derivatives
Full size table

In vitro PPPA metabolism in human-, rabbit-, rat- and mouse-liver microsomes

PPPA (1) was incubated with human-, rabbit-, rat- and mouse-liver microsomes in the NADPH-generating system for various incubation periods. Without liver microsomes, 1 was stable and no PPPA metabolites were observed (data not shown). With liver microsomes, five metabolites (M1∼M5) were observed from human in UFLC analysis, three metabolites (M1, M2 and M5) were from rabbit, three metabolites (M1, M3 and M4) were from rat and three metabolites (M1, M2 and M3) were from mouse (Figure 2). As shown in Table 1, the retention times of M1, M3, M4 and M5 were found to be 2.8, 3.1, 3.5 and 3.6 min, respectively, suggesting that they were 1,11-dideacetyl PPPA (6), 1-deacetyl PPPA (2), 11-deacetyl PPPA (4) and N-oxidized PPPA (10), respectively. Furthermore, their [M+H]+ were detected as 500, 542, 542 and 600 by LC-MS, confirming the structure as expected. However, the structure of M2 was not defined.

Figure 2
figure 2

Representative ultra fast liquid chromatogram after incubation of pyripyropene A (PPPA) (1) in human, rabbit, rat and mouse-liver microsomes. PPPA (1) was incubated for 0, 30, 120, 360 or 540 min in the presence of human, rabbit, rat or mouse liver microsomes. The metabolites were extracted with ethyl acetate and analyzed by ultra fast liquid chromatography under the following conditions: column, Shim Pack XR-ODS (Shimadzu), 2.0 × 75 mm; solvent, 6-min linear gradient from 5% acetonitrile in 0.1% phosphoric acid to 95% acetonitrile in 0.1% phosphoric acid; flow rate, 0.55 ml min−1; detection, UV radiation at 320 nm.

Full size image

PPPA (1) was rapidly metabolized to M1 and M2 in mouse-liver microsomes, and to M1 in rat and rabbit liver microsomes, while slowly to M3 in human-liver microsomes. Furthermore, the half life times of 1 were found to be 520, 55, 72 and 15 min in human, rabbit, rat and mouse-liver microsomes, respectively (Figure 3).

Figure 3
figure 3

Metabolism of pyripyropene A (PPPA) (1) in human-, rabbit-, rat- and mouse-liver microsomes. PPPA (1) was incubated in liver microsomes at 37 °C, and production of 1 and M1 to M5 were measured and calculated as peak area of PPPA metabolite (%). The initial peak area of 1 was defined as 100%. Each value represents the mean±s.e.m. of more than three independent experiments.

Full size image

Next, the structure of M2 was elucidated as follows: M2 was purified with high-performance liquid chromatography (column, YMC-Pack, 6 × 250 mm (YMC, Kyoto, Japan); solvent, a 20-min gradient from 20% acetonitrile to 40% acetonitrile containing 0.05% trifluoroacetic acid; flow rate, 8.0 ml min−1; detection, UV radiation at 210 nm), and analyzed by the linked-scan mode of fast atom bombardment mass spectroscopy (FAB-MS: JMS-DX300 (JEOL, Tokyo, Japan); Matrix: thioglycerol). As shown in Figure 4, the fragment ion peaks were observed at m/z 79, 91, 104, 148, 202, 218, 258, 384, 392, 454 and 468 in FAB-MS, suggesting that M2 was 1-deacetyl-11-deacetoxymethyl-11-formyl PPPA (9). Furthermore, the proposed derivative 9 was synthesized, confirming that all the UFLC and LC-MS data of M2 were identical with those of 9 (Table 1). Thus, M2 was elucidated to be 9.

Figure 4
figure 4

Mass fragments analysis of pyripyropene A metabolite M2.

Full size image

In vitro PPPA metabolism in human, rabbit, rat and mouse plasma

PPPA (1) was incubated with human, rabbit, rat and mouse plasma for various incubation periods. PPPA (1) was stable in human, rabbit and rat plasma at least within 540 min (Figure 5), while 1 was slowly metabolized to M6 in mouse plasma. M6 was eluted at 3.0 min by UFLC and at 7.4 min with m/z [M+H]+ of 542 by LC-MS (Table 1), indicating that M6 is 7-deacetyl PPPA (3).

Figure 5
figure 5

Representative ultra fast liquid chromatogram after incubation of pyripyropene A (PPPA) (1) in human, rabbit, rat and mouse plasmas. PPPA (1) was incubated for 0, 30, 120, 360 or 540 min in the presence of human, rabbit, rat or mouse plasma. The metabolites were extracted with ethyl acetate and analyzed by ultra fast liquid chromatography under the following conditions: column, Shim Pack XR-ODS (Shimadzu), 2.0 × 75 mm; solvent, 6-min linear gradient from 5% acetonitrile in 0.1% phosphoric acid to 95% acetonitrile in 0.1% phosphoric acid; flow rate, 0.55 ml min−1; detection, UV radiation at 320 nm.

Full size image

Effect of PPPA metabolites on ACAT isozymes

It is important to see whether or not PPPA metabolites inhibit ACAT isozymes. Therefore, synthetic PPPA derivatives were tested in the cell-based assay using ACAT1- and ACAT2-Chinese hamster ovary cells. As shown in Table 2, all derivatives lost or markedly decreased ACAT2 inhibitory activity as compared with 1. 1-deacetyl and/or 11-deacetyl derivatives (2, 4 and 6) still showed weak inhibition but were 170 to 500 times less potent than 1.

Table 2 Effect of PPPA and metabolites on ACAT1 and ACAT2 isozymes
Full size table

Discussion

PPPA (1), a potent and selective inhibitor of ACAT2, recently proved orally active in atherogenic mouse models.25 Since 1 was originally discovered in 1993, new PPPA derivatives (the second generation derivatives) with more potent and more selective activity against ACAT2 than 1 were required. For this purpose, potential in vitro metabolism of 1 was investigated.

From the present data, the in vitro metabolic pathway of 1 in the liver and plasma was illustrated in Scheme 3. In mouse liver (Figure 3), 1 was exclusively metabolized to M3 (2) then to M1 (6), indicating that hydrolysis of 1 occurs first at the 1-O-acetyl residue then at the 11-O-acetyl residue. The 7-O-acetyl residue was found to be resistant to hydrolytic enzymes in the mouse-liver microsomes. This ordered metabolic pathway of 1 appeared common in the liver of all species tested, although the half life times of 1 are distinct among the species; mouse liver with the shortest half life of 15 min and human liver with the longest one of 520 min, indicating that the esterase activity in human is lower than that in nonhuman primates and rodents. These findings were comparable with the previous studies.27, 28, 29 Interestingly, there are minor metabolic pathways of 1 in human liver; first hydrolysis occurred at the 11-O-acetyl residue to produce 4 and N-oxidation occurred to produce 10. In human, rat and rabbit plasma, 1 was stable, but in mouse plasma 1 was hydrolyzed at 7-O-acetyl residue.

From the in vitro metabolic study of 1, there are two issues to be made clear. First, it was demonstrated that 1 is more stable in human than in mouse, suggesting that the doses for human use might be lower (<25 mg/kg/day) than those for mouse use (doses >25 mg/kg/day are effective to show atheroprotective activity in mouse models25). Second, 1 is mainly hydrolyzed at the 1-O-acetyl and the 11-O-acetyl residues, while the 7-O-acetyl residue is resistant. These PPPA metabolites had much lower ACAT inhibitory activity compared with 1. Based on these findings, the second generation PPPA derivatives were semi-synthetically prepared, leading to the discovery of promising candidates, which are resistant to hydrolytic enzymes in liver microsomes.1, 30, 31, 32 Furthermore, we found that certain PPPA derivatives proved more potent in atherogenic mouse and rabbit models than 1. The detail will be published elsewhere.

In summary, we showed in vitro the main metabolic pathways of 1 in liver microsome and plasma of human, rabbit, rat and mouse. Our findings are expected to support the synthesis studies of PPPA derivatives, safety studies and clinical trials for the development of novel anti-atherosclerotic drugs.

Experimental procedure

Materials

Acetonitrile (high-performance liquid chromatography grade) and phosphoric acid were obtained from Kanto Chemicals (Tokyo, Japan). D-Glucose-6-phosphate and glucose 6-phosphate dehydrogenase were from Wako Pure Chemicals (Osaka, Japan). [1-14C]Oleic acid (1.85 GBq mmol) was from PerkinElmer (Waltham, MA, USA). Ham's nutrient mixture, F-12 medium, trifluoroacetic acid and NADP were from Sigma Aldrich (St Louis, MO, USA). Penicillin (10 000 units ml−1)/streptomycin (10 000 μg ml−1) solution and MEM vitamins were from Invitrogen (Carlsbad, CA, USA). Geneticin was from Calbiochem (Darmstadt, Germany). Fetal bovine serum was from Hyclone (Waltham, MA, USA). Human-liver microsomes (pooled from 50 donors of mixed gender) and liver microsomes from New Zealand rabbits (male), Sprague Dawley rats (female) and CD-1 mice (female) were purchased from Xenotech (Lenexa, KS, USA). Plasma from humans (pooled from 50 donors) was obtained from George King Bio-Medical (Overland Park, KS, USA), while plasma from rabbits (New Zealand rabbit of mixed sex) and rat (Wistar rat of mixed sex) was from Rockland (Gilbertsville, PA, USA). PPPA (1) was purified from the culture broth of A. fumigatus FO-1289 according to our established method.15, 16, 17 Blood of female C57BL/6J mice (female, Japan SLC, Shizuoka, Japan) was collected in a container containing sodium EDTA and centrifuged for 10 min at 3000 r.p.m. at 4 °C. The plasma was frozen and stored at −80 °C. Mice were maintained and cared for according to the regulations of the Animal Care Committees of Kitasato University.

Synthesis of pyripyropene derivatives

Possible PPPA derivatives (reference standards) were synthesized according to the previously reported methods.21, 22, 23 Known PPPA derivatives 3, 8 and 10 (Table 1) were synthesized according to the previously reported methods from 1.21, 22, 23 The other novel PPPA derivatives 2, 4, 5, 6, 7 and 9 (Table 1) were prepared on the basis of the synthetic strategy developed by our group.31, 33 1H- and 13C-NMR spectra were obtained on VARIAN MERCURY plus-300 and XL-400 (Varian Inc., Palo Alto, CA, USA), and chemical shifts were reported on the δ scale and referenced to tetramethylsilane. Mass spectra were measured with JMS-AX505HA spectrometers (JEOL, Tokyo, Japan). Unless otherwise indicated, commercial reagents were used without further purification. Organic solvents were distilled and dried over molecular sieves (3 or 4 Å) prior to use. Reactions were carried out in flame-dried glassware under a positive pressure of argon with magnetic stirring, unless otherwise indicated. Flash column chromatography was carried out on silica gel 60N (spherical, neutral, particle size: 40–50 μm). Thin layer chromatography was performed on 0.25-mm silica gel 60 F254 plates (Merck, Darmstadt, Germany). Plates were visualized by UV radiation (254 nm) and anisaldehyde staining. The detail of these syntheses is described below.

Assay for in vitro metabolism in liver microsomes or plasma

Experiments of in vitro metabolism of 1 in liver microsomes and plasma were carried out according to the established method34 with minor modifications. Microsomes (0.20 pmol P450 ml−1) or plasma (10% (v/v)) were incubated with 1 (6.0 μg ml−1 in the incubation buffer containing 100 mM sodium phosphate buffer, pH 7.4, 1.0 mM EDTA, 5.0 mM D-glucose-6-phosphate, 1.0 mM NADP and 1.0 unit ml−1 glucose 6-phosphate dehydrogenase) in a total volume of 200 μl at 37 °C. The incubation was terminated by adding ethyl acetate (500 μl) and distilled water (300 μl), which were mixed well. After centrifugation to separate the ethyl acetate and water layers, ethyl acetate layer (300 μl) was evaporated to dryness in vacuo.

UFLC analysis

For quantitative analysis of PPPA metabolites, the samples dissolved in methanol were analyzed by UFLC (Prominence, Shimadzu, Kyoto, Japan) under the following conditions: column, Shim Pack XR-ODS (Shimadzu), 2.0 × 75 mm; column temperature, 50 °C; solvent, 6-min linear gradient from 5.0% acetonitrile in 0.10% phosphoric acid to 95% acetonitrile in 0.10% phosphoric acid; flow rate, 0.55 ml min−1; detection, UV radiation at 320 nm.

LC-MS analysis

For structural information of PPPA metabolites, the samples dissolved in methanol were analyzed by an LC-MS system. PPPA metabolites were separated by high-performance liquid chromatography (Agilent 1200 Series LC system) under the following condition: column, Pegasil ODS (Senshu Sci. Corp., Tokyo, Japan), 2 × 150 mm; column temperature, 25 °C; solvent, 12-min linear gradient from 10% acetonitrile in 0.10% formic acid to 100% acetonitrile in 0.10% formic acid; flow rate, 0.20 ml min−1; detection, UV radiation at 320 nm. After each peak was separated by high-performance liquid chromatography, they were analyzed by AccuTOF time-of-flight mass spectrometer (JEOL).

Assay for [14C]cholesteryl ester synthesis in ACAT1- or ACAT2-Chinese hamster ovary cells

The assay for the synthesis of [14C]cholesteryl ester from [14C]oleic acid in ACAT1- or ACAT2-expressing Chinese hamster ovary cells18 was carried out by our established method.19 A stock solution of PPPA and its derivatives was made to a concentration of 10 mg ml−1 in methanol. In brief, both cell lines were maintained at 37 °C in 5.0% CO2 in Ham’s F-12 medium supplemented with MEM vitamins, geneticin (G418 sulfate) (300 μg ml−1) and 10% heat-inactivated fetal bovine serum (hereafter referred to as medium A). ACAT1- or ACAT2-Chinese hamster ovary cells (1.25 × 105 cells per 0.25 ml in medium A) were cultured in a 48-well plastic microplate and allowed to recover overnight at 37 °C in 5.0% CO2. The assay was done with cells that were at least 80% confluent. Following the overnight recovery, a sample (2.5 μl in methanol solution) over the final concentration range of 0.0001–10 μg ml−1 with two wells for each concentration and [1-14C]oleic acid (5.0 μl in 10% ethanol/phosphate buffered saline solution, 1.0 nmol, 1.85 KBq) were added to each culture at 37 °C in 5.0% CO2. After a 6-h incubation, the medium was removed, and the cells in each well were washed twice with phosphate buffered saline. The cells were lysed by adding phosphate buffered saline (0.25 ml) containing 0.10% (w/v) sodium dodecyl sulfate, and the cellular lipids were extracted by the method of Bligh and Dyer.35 After concentrating the organic solvent, the total lipids were separated on a TLC plate (silica gel F254, 0.5-mm thick, Merck, Darmstadt, Germany) and analyzed with a BAS2000 (Fuji Film, Tokyo, Japan). In this assay, [14C]cholesteryl ester was produced by the reaction of ACAT1 or ACAT2. ACAT inhibitory activity (%) is defined as (1−[14C]cholesteryl ester-drug/[14C]cholesteryl ester-control) × 100. The IC50 value is defined as the drug concentration causing 50% inhibition of a biological activity, and is calculated from the curve fit of the data using a sigmoidal dose-response curve.

scheme 1

Synthesis of novel pyripyropene A derivatives 2, 4, 6 and 9. Reagents and conditions: (a) (tBu)2Si(OTf)2, 2,6-lutidine, DMF (dimethylformamide), r.t., 0.5 h, quant.; (b) Ac2O, Et3N, DMAP, CH2Cl2, 0 °C, 1 h, quant.; (c) NH4F, MeOH, r.t., 3 h, 13: 78%, 14: 5%, 6: 5%; (d) (i) TEMPO, IBDA, CH2Cl2, r.t., 2 h (ii) Et3N 3HF, THF, r.t., 0.5 h, two steps, 47%; (e) (i) Ac2O, Et3N, DMAP, CH2Cl2, 0 °C, 1 h (ii) Et3N 3HF, THF, r.t., 0.5 h, 2: two steps, 90%, 4: two steps, 84%.

scheme 2

Synthesis of novel pyripyropene A derivatives 5 and 7. Reagents and conditions: (a) Alloc-Cl, TMEDA, CH2Cl2, 0 °C, 1 h, 67%; (b) NH4F, MeOH, r.t., 3 h, 16: 89%, 17: 5%; (c) (i) Ac2O, Et3N, DMAP, CH2Cl2, 0 °C, 1 h (ii) Pd(PPh3)4, HCO2NH4, THF, r.t., 1 h (iii) Et3N 3HF, THF, r.t., 0.5 h, 5: three steps, 73%, 7: three steps, 57%.

scheme 3

Proposed metabolic pathway for pyripyropene A (PPPA) (1). The line indicated the metabolized potency: bold line>line>broken line.