Differential Reactivity of Purified Bioactive Coffee Furans, Cafestol and

Nov 6, 2009 - di Chimica Biomolecolare del CNR, Comprensorio OliVetti, Edificio A, Via Campi Flegrei 34,. I-80078 Pozzuoli Naples, Italy. ReceiVed Jul...
0 downloads 0 Views 256KB Size
Download PDF
1922

Chem. Res. Toxicol. 2009, 22, 1922–1928

Differential Reactivity of Purified Bioactive Coffee Furans, Cafestol and Kahweol, with Acidic Nitrite: Product Characterization and Factors Controlling Nitrosation Versus Ring-Opening Pathways Maria De Lucia,† Lucia Panzella,† Dominique Melck,‡ Italo Giudicianni,§ Andrea Motta,‡ Alessandra Napolitano,*,† and Marco d’Ischia† Department of Organic Chemistry and Biochemistry and Centro Interdipartimentale di Metodologie Chimico-Fisiche (CIMCF), UniVersity of Naples “Federico II”, Via Cinthia 4, I-80126 Naples, Italy, and Istituto di Chimica Biomolecolare del CNR, Comprensorio OliVetti, Edificio A, Via Campi Flegrei 34, I-80078 Pozzuoli Naples, Italy ReceiVed July 3, 2009

Cafestol and kahweol, coffee-specific furan diterpenes, are believed to cause various physiological effects in human subjects, including an increase in cholesterol and plasma triacylglycerol levels as well as cancer chemopreventive effects. Despite the increasing interest in these compounds raised by the diverse range of biological activities, their reaction behavior and degradation pathways under physiologically relevant conditions remain uncharted. Herein, we report a detailed investigation of the structural modifications suffered by cafestol and kahweol in the presence of acidic nitrite under conditions mimicking those occurring in the stomach during digestion as well as by action of other oxidants. Prior to the chemical study, an isolation procedure for kahweol from green coffee beans was developed based on Soxhlet extraction followed by preparative HPLC. Preliminary experiments showed that kahweol is much more reactive than cafestol toward nitrite at pH 3, as evidenced by inhibition experiments with the 2,3-diaminonaphthalene assay as well as by product analysis in coffee extracts. When exposed to equimolar nitrite in phosphate buffer, pH 3, kahweol gave as a main product the ring-opened dicarbonyl derivative 1. Under more forcing conditions, cafestol reacted as well to give a main nitrogenous product identified as the 1-hydroxy-2-pyrrolinone 2. It is concluded that the conjugated double bond in kahweol is a critical structural element, increasing the susceptibility of the furan ring to protonation rather than nitrosation and favoring ring-opening routes driven by the irreversible oxidation steps. These results offer a useful background to assess the effects of coffee-specific lipids in association with abnormally high nitrite levels from the diet. Introduction Cafestol and kahweol (Chart 1) are pentacyclic kaurane diterpenes specifically found in coffee beans and brews. They share as a common structural feature a disubstituted furan ring and differ only in the presence (kahweol) or absence (cafestol) of a double bond conjugated to the furan ring. Cafestol is contained in both Coffea arabica and Coffea canephora (Robusta), whereas kahweol is mainly present in Arabica, while a cafestol derivative, 16-O-methylcafestol, is found only in Robusta and is accepted as a specific marker (1). In coffee oil, the diterpenes occur in the form of esters with fatty acids. The contents of cafestol esters vary between 2.0 and 8.0 g/kg dry weight in Robusta coffee and 9.0 and 22.0 g/kg dry weight in Arabica coffee, and roasting appears to have little influence on their percentage composition in coffee brew (2). Because of the wide consumption of coffee in the world, considerable interest has been focused during the past decades on the potential effects of its main specific constituents to human health and disease incidence. Extensive epidemiological studies * To whom correspondence should be addressed. Tel: +39081674133. Fax: +39081674393. E-mail: [email protected]. † Department of Organic Chemistry and Biochemistry, University of Naples “Federico II”. ‡ Istituto di Chimica Biomolecolare del CNR. § Centro Interdipartimentale di Metodologie Chimico-Fisiche (CIMCF), University of Naples “Federico II”.

Chart 1

indicated in particular that consumption of boiled coffee is associated with an increase in the plasma levels of cholesterol, which has been attributed to the stimulating effects of furan diterpenes, especially cafestol, regarded as one of the most potent cholesterol-elevating compounds known in the human diet (3, 4). Several genes involved in cholesterol homeostasis have been shown to be targets of cafestol, including notably those encoding for cholesterol 7R-hydroxylase (CYP7A1), the rate-limiting enzyme in bile acid biosynthesis, and it has been shown that cafestol elevates serum lipid levels by acting as an agonist ligand for nuclear hormone receptors farnesoid X receptor and pregnane X receptor, thus affecting cholesterol homeostasis (3). In other studies, cafestol has also been reported to raise plasma triacylglycerol levels by increasing the production rate of VLDL1 apoprotein B, probably via increased assembly of VLDL1 in the liver (5). Beside the significant and well-documented hyperlipidemic activity, coffee furan diterpenes have been associated with

10.1021/tx900224x CCC: $40.75  2009 American Chemical Society Published on Web 11/06/2009

Reaction of Cafestol and Kahweol with Nitrous Acid

diverse beneficial health effects, including a chemopreventive activity against liver and colon cancer by inhibiting the mutagenicity/tumorigenicity of several carcinogens in animal models and cell culture systems (6, 7). Protective effects have been demonstrated against 7,12-dimethylbenz[a]anthracene, aflatoxin B1 (8), benzo[a]pyrene (9), N-nitrosodimethylamine, and 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (10). Other activities of kahweol and cafestol concern suppression of COX-2 expression, prostaglandin E2, and COX-2 protein production in lipopolysaccharide-activated macrophages (11). Cafestol moreover blocks inducible nitric oxide synthase expression (12), whereas kahweol has anti-inflammatory and antiatherosclerotic activities, which involve inhibition of tumor necrosis factor-R-induced expression of cell adhesion molecules in human endothelial cells (13). Finally, both diterpenes have been shown to be antioxidants, protecting against H2O2-induced oxidative stress and DNA damage, probably via scavenging free oxygen radicals (14). Despite the increasing interest in bioactive coffee-specific diterpenes, studies of the reaction behavior and possible modifications of these compounds under conditions of physiological relevance are surprisingly lacking. Moreover, in most of the studies, data refer to cafestol or mixtures of cafestol and kahweol, whereas results with pure kahweol are not available due to difficulties with purification and stability, hampering a more in-depth investigation of this specific coffee constituent. A most interesting issue relates to the possible transformations undergone by cafestol and kahweol in the acidic environment of the stomach following interaction with specific dietary and saliva constituents, notably nitrite. Exposure to excess nitrite from the diet, for example polluted drinking water, spinaches and vegetables, or cured/pickled meats, or from aberrant endogenous generation of nitric oxide at sites of chronic inflammation is implicated as a potential etiological factor in the development of stomach and colorectal cancers (15, 16). In the acidic environment of the stomach (pH 2.5-4.0 during digestion), nitrite is in equilibrium with nitrous acid (pKa ) 3.2) (17), which is unstable and decomposes to a range of reactive nitrogen species (RNS),1 such as nitrogen dioxide and the nitrosonium ion. The pathological consequences associated with excessive nitrous acid accumulation and aberrant RNS formation in the stomach are rooted in their ability to cause a series of toxic events such as (1) DNA base deamination by diazotization of primary amino groups and mutagenesis (18); (2) carcinogenic nitrosamine formation (19, 20); (3) nitration of biomolecules, including chiefly tyrosine residues in proteins but also unsaturated fatty acids and nucleic acids (21, 22); and (4) methemoglobin formation, with a consequent decrease in oxygen availability and tissue asphyxia (23). That nitrite-derived RNS may cause severe structural modifications to furan-containing compounds has recently been demonstrated by a study on menthofuran, a toxic terpenoid from various mint oils (24). In the present study, we report for the first time a chemical insight into the acid-promoted reactions of cafestol and kahweol with nitrite ions under conditions aimed to model interactions that may occur in the gastric compartment following elevated nitrite intake (25). Specific aims were to isolate and characterize the main products, to elucidate the reaction mechanisms and the underlying 1 Abbreviations: RNS, reactive nitrogen species; DAN, 2,3-diaminonaphthalene; CAN, ceric ammonium nitrate; COSY, correlation spectroscopy; TOCSY, total correlation spectroscopy; ROESY, rotating frame Overhauser enhancement spectroscopy; HR, high resolution; ESI+, electrospray ionization in positive ion mode.

Chem. Res. Toxicol., Vol. 22, No. 12, 2009 1923

role of structural factors, and to assess whether and to what extent this chemistry may be of toxicological relevance, for example, in nitrite scavenging or in mutagen generation.

Experimental Procedures Materials and Methods. Green or roasted C. arabica beans from Colombia were used. Roasting was carried out in an oven at 230-240 °C for 20 min. Caffeic acid, 2,3-diaminonaphthalene (DAN), 2′-deoxyguanosine, 2′-deoxyguanosine 5′-monophosphate, sodium nitrite, ammonium iron(II) sulfate hexahydrate, EDTA, ceric ammonium nitrate (CAN), and hydrogen peroxide (30% solution) were used as obtained. NMR spectra were recorded at 600.13 MHz on a Bruker DRX-600 spectrometer, equipped with a TCI CryoProbe, fitted with a gradient along the Z-axis. Homonuclear correlation spectroscopy (COSY), total correlation spectroscopy (TOCSY), and rotating frame Overhauser enhancement spectroscopy (ROESY) spectra were recorded at 300 K by using standard Bruker pulse programs with the time proportional phase incrementation of the first pulse. Chemical shifts are reported in δ values downfield from TMS. High resolution (HR) electrospray ionization in positive ion mode (ESI+)/MS spectra were obtained in 2% formic acid-methanol 1:1 v/v. Analytical HPLC was carried out on an instrument equipped with an UV detector set at 220 or 254 nm. The chromatographic separation was achieved on an octadecylsilane coated column, 250 mm × 4.6 mm, 5 µm (Sphereclone, Phenomenex), using binary gradient elution conditions at a flow rate of 0.7 mL/min as follows: water, solvent A; acetonitrile, solvent B; from 25 to 45% B, 0-20 min; 45% B, 20-40 min; from 45 to 65% B, 40-50 min (eluant I); water, solvent A; acetonitrile, solvent B; 10% B, 0-20 min; from 10 to 45% B, 20-40 min; 45% B, 40-60 min; from 45 to 65% B, 60-70 min (eluant II); 20 mM ammonium formate (pH 4.0), solvent A; acetonitrile, solvent B; from 10 to 80% B, 0-50 min (eluant III); water, solvent A; acetonitrile, solvent B; from 2 to 65% B, 0-50 min (eluant IV). For preparative purposes, HPLC was carried out on an instrument equipped with an UV detector set at 220 or 254 nm, and the chromatographic separation was achieved on a preparative Econosil C18 column (250 mm × 22 mm, 10 µm) or a semipreparative Econosil C18 column (250 mm × 10 mm, 10 µm). The purity of isolated compounds was estimated by 1H NMR analysis. Isolation of Cafestol and Kahweol from Coffee Beans. Green coffee beans were ground in liquid nitrogen using an electric homogenizer. Ground coffee beans (60 g) were then extracted with a Soxhlet apparatus with n-hexane (600 mL) for 18 h at 80 °C. The oil thus obtained (8 g, 13% w/w) was treated under an argon atmosphere with 5% KOH (5 mL) in methanol under stirring at room temperature. After 2 h, the reaction mixture was diluted with n-hexane and extracted three times with methanol/water 9:1 v/v. After water removal by use of a rotatory evaporator, the methanolic phase was treated again with KOH (400 mg) under the same conditions as above for 30 min. The reaction mixture was then diluted with water and extracted with dichloromethane containing 8% methanol v/v. The residue obtained from the organic phase after removal of the solvent (1.45 g, 18% yield with respect to the starting oil) was dissolved in methanol (0.2 g/mL final concentration) and fractionated by preparative HPLC (eluant water/acetonitrile 45:55 v/v, flow rate 20 mL/min) to give kahweol (tr 30 min, 46 mg, purity > 98%) and cafestol (tr 32 min, 22 mg, purity > 98%). DAN Assay. The reaction was carried out as described (26) with slight modifications. To a solution of cafestol, kahweol, or caffeic acid (0-0.2 × 10-3 mmol) and DAN (0.04 × 10-3 mmol) in methanol (60 µL), 50 mM sodium acetate buffer (pH 4.0, 140 µL) was added, followed by sodium nitrite (0.004 mmol). After 30 min, 50 mM sodium phosphate buffer (pH 7.4, 1.8 mL) was added to stop the reaction. Naphtho[2,3-d]triazole was quantified by measuring the fluorescence of each sample using an excitation wavelength of 375 nm and an emission wavelength of 450 nm. In other experiments, to a solution of caffeic acid, kahweol, and cafestol (1 × 10-3 mmol) in methanol (300 µL), 50 mM sodium acetate buffer (pH 4.0, 700 µL) was added, followed by sodium nitrite (0.02

1924

Chem. Res. Toxicol., Vol. 22, No. 12, 2009

mmol). The reaction mixture was taken under stirring at room temperature and periodically analyzed by HPLC (eluant II). DAN assay was also carried out on product 1 and 2 (see below) under the same conditions adopted for cafestol and kawheol. Reaction of Coffee with Nitrite Ions. Coffee beverage was prepared by infusion in boiling distilled water (150 mL) for 10 min of coffee powder (9 g) obtained from finely ground roasted coffee beans. The infusion was left at room temperature for 30 min, and the powder that was deposited was removed by gentle decantation without filtering. The coffee infusion (100 mL) was acidified to pH 3 with 3 M HCl and treated with 2 mM NaNO2. After 2 h, the mixture was extracted with ethyl acetate (3 × 100 mL), and the combined organic layers were dried over sodium sulfate and taken to dryness. The residue obtained was dissolved in methanol (1 mL) and analyzed by HPLC (eluant I). In another series of experiments, the reaction with nitrite ions was run on coffee oil obtained from green coffee beans and subjected to KOH hydrolysis as described above. A 1 mM concentration of NaNO2 was added to a solution of the hydrolyzed coffee oil in 0.1 M sodium phosphate buffer (pH 3.0) containing cafestol and kawheol at 1 mM overall concentration. The reaction mixture was taken under stirring at room temperature and periodically analyzed by HPLC (eluant I). Reaction of Cafestol and Kahweol with Nitrite Ions. To 3 mg of cafestol or kahweol dissolved in methanol (2 mL), 8 mL of 0.1 M sodium phosphate buffer (pH 3.0) was added (1 mM final concentration) followed by NaNO2 (1 mM). The reaction mixture was taken under stirring at room temperature and periodically analyzed by HPLC (eluant I). In other experiments, the reaction of cafestol or kahweol was run (i) in the absence of nitrite, (ii) under an argon atmosphere, (iii) using a substrate concentration in the range of 0.1-1 mM, and (iv) with nitrite varying in the range of 1-10 mol equiv. Isolation of 1. For preparative purposes, the reaction of kahweol (1 mM) with NaNO2 (1 mM) at pH 3 was run on 50 mg of the starting material previously dissolved in methanol (30 mL). After 2 h, the reaction mixture was extracted with ethyl acetate (3 × 50 mL), and the combined organic layers were dried over sodium sulfate and taken to dryness. The residue was fractionated by semipreparative HPLC (eluant water/acetonitrile 75:25 v/v, flow rate 4 mL/min) to give 1 (tr 15 min with eluant I, 12 mg, 23% yield, purity > 90%). Compound 1: HR ESI+/MS: m/z 331.1918, calcd for C20H27O4, 331.1909 ([M + H]+); for 1H and 13C NMR spectra and positional assignment, see the Supporting Information and Table 2. Isolation of 2. For preparative purposes, the reaction of cafestol (1 mM) with NaNO2 (10 mM) at pH 3 was run on 30 mg of the starting material previously dissolved in methanol (20 mL). After 2 h, the reaction mixture was extracted with ethyl acetate (3 × 50 mL), and the combined organic layers were dried over sodium sulfate and taken to dryness. The residue was fractionated by semipreparative HPLC (eluant water/acetonitrile 75:25 v/v, flow rate 4 mL/min) to give 2 (tr 6 min with eluant I, 5 mg, 15% yield, purity > 95%). Compound 2: HR ESI+/MS: m/z 364.2135, calcd for C20H30NO5, 364.2124 ([M + H]+); for 1H and 13C NMR spectra and positional assignment, see the Supporting Information and Table 2. Reaction of Kahweol with the Fenton Reagent. To 3 mg of kahweol dissolved in methanol (2 mL), 8 mL of 0.1 M sodium phosphate buffer (pH 7.4) were added (1 mM final concentration) followed by 1 mol equiv of a complex Fe(NH4)2(SO4)2 × 6H2O/ EDTA (1:1 molar ratio) and 1 mol equiv of H2O2. The reaction mixture was taken under vigorous stirring at room temperature and periodically analyzed by HPLC (eluant I). Reaction of Kahweol with CAN. To 3 mg of kahweol dissolved in methanol (2 mL), 8 mL of 0.1 M sodium phosphate buffer (pH 3.0) were added (1 mM final concentration) followed by 1 mol equiv of CAN. The reaction mixture was taken under vigorous stirring at room temperature and periodically analyzed by HPLC (eluant I). Reaction of 1 with 2′-Deoxyguanosine. The reaction was run under conditions previously reported (27). Briefly, 1 (0.06 mM) was incubated with 2′-deoxyguanosine 5′-monophosphate (1 mg/

De Lucia et al. Table 1. 1H and 13C NMR Data of Kahweola kahweol H (mult, J)b

1 c

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 C10-CH3 b

5.98 (d, 10.2) 6.24 (d, 10.2) 2.60 (dd, 12.6, 3.0) 1.70 (m), 1.93 (m) 1.65 (m), 1.80 (m) 1.62 (m) 1.75 1.55 2.10 1.70 1.50

(m), (m), (m) (m), (m),

1.85 (m) 1.75 (m)

3.67 6.37 7.35 1.01

(d, 11.4), 3.78 (d, 11.4) (d, 1.2) (d, 1.2) (s)

1.95 (m) 1.65 (m)

13

C

139.7 116.2 151.5 123.1 45.8 23.1 41.7 46.1 49.8 43.1 20.1 26.7 46.3 39.1 53.8 82.9 66.9 109.6 142.5 16.1

a Spectra run in CD3OD; chemical shift values are given in ppm. J expressed in Hz. c Numbering as shown in Chart 1.

mL) in 0.1 M phosphate buffer (pH 4.0). The reaction mixture was taken under stirring and periodically analyzed over 24 h by HPLC (eluant III). In other experiments, the reaction was run using a 10:1 substrate to 2′-deoxyguanosine ratio as described (28) but with 1 at 33 mM and 2′-deoxyguanosine at 3 mM in 0.1 M phosphate buffer (pH 4.0 or 7.0). The reaction course was monitored by HPLC (eluant IV).

Results and Discussion Kahweol Purification and Spectral Characterization. Kahweol is not easily accessible in pure form in more than small amounts because of the considerable difficulties in separation from cafestol due to the marked structural similarity. Whereas cafestol is commercially available as the acetate ester, requiring hydrolysis prior to use, or can be prepared by reduction of mixtures of the two compounds from coffee bean extracts (29, 30), the only known isolation procedure for kahweol relies on a laborious chromatographic fractionation on silver impregnated silica gel (31). Accordingly, to permit chemical studies, a purification procedure of kahweol from coffee bean extracts was initially developed. Typically, ground green coffee beans were extracted with a Soxhlet apparatus with n-hexane for 18 h at 80 °C. The oil thus obtained (8 g, 13% w/w) was treated with 5% KOH/CH3OH to achieve fatty acid-ester transesterification. Eventually, after a further hydrolytic treatment, a residue consisting mainly of cafestol and kahweol was obtained. This was subjected to preparative HPLC on a reverse phase column (acetonitrile/water 55:45 v/v). The isolated compounds were pure as determined by analytical HPLC (see the Supporting Information) and NMR analysis. A complete characterization of kahweol by 2D NMR techniques was also performed (Table 1). Reactivity of Cafestol and Kahweol toward Acidic Nitrite in Coffee. To assess the possible consequences of acidic nitrite exposure on cafestol and kahweol in coffee, an unfiltered brew was prepared by infusion of the finely ground roasted beans in boiling water for 10 min. This specific brew preparation procedure was chosen to maintain the diterpene levels expectedly found in coffee drinks commonly served in many European and Mid-Eastern countries (32). The coffee brew was thus treated with nitrite ions up to 2 mM concentration at pH 3, and cafestol/kahweol levels were

Reaction of Cafestol and Kahweol with Nitrous Acid

Figure 1. HPLC elution profiles of the reaction mixture of coffee oil after the hydrolysis treatment containing kahweol and cafestol in a 1:1 ratio (1 mM overall concentration) with nitrite ions (1 mM) at pH 3 (trace a, 0 min; trace b, 1 h). Elution conditions are described in the Experimental Procedures. Detection at 220 nm.

Chem. Res. Toxicol., Vol. 22, No. 12, 2009 1925

Figure 3. HPLC elution profiles of the reaction mixture of caffeic acid, kahweol, and cafestol (1 mM each) with nitrite ions (20 mM) at pH 4 (trace a, 0 min; trace b, 30 min). Elution conditions are described in the Experimental Procedures. Detection at 220 nm.

Figure 4. HPLC elution profiles of the reaction mixture of kahweol (1 mM) with nitrite ions (1 mM) at pH 3 (trace a, 0 min; trace b, 2 h). Elution conditions are described in the Experimental Procedures. Detection at 220 nm.

Figure 2. Inhibition of N-nitrosation of DAN by cafestol ([), kahweol (9), and caffeic acid (2) measured as fluorescence emission at 450 nm of naphtho[2,3-d]triazole. Relative fluorescence represents the ratio of fluorescence values measured in the presence and in the absence of the inhibitor. Shown are the mean ( SD values for two separate experiments (n ) 3).

determined after 2 h by HPLC analysis of the beverage following ethyl acetate extraction against a control mixture in the absence of nitrite. The HPLC profile (Supporting Information) of the organic extract clearly showed a marked decrease in both compounds, with kahweol being significantly more reactive than cafestol. Because both compounds occur mostly in the ester form, in further experiments, the reactivity of kahweol and cafestol toward acidic nitrite was also explored using coffee oil previously subjected to a hydrolytic treatment. Figure 1 shows the HPLC profile of the coffee oil, in which the compounds are present at about a 1:1 ratio, prior and following treatment with equimolar nitrite in phosphate buffer, pH 3. A faster consumption of kahweol (ca. 85% after 1 h) relative to cafestol (ca. 25% after 1 h) was observed with the formation of more polar species that eluted earlier than the starting furans under the elutographic conditions chosen. In subsequent experiments, the activity of cafestol and kahweol as inhibitors of nitrosation reactions was investigated with the DAN assay (26) using the coffee polyphenol caffeic acid as a reference compound. Measurement of fluorescent naphtho[2,3-d]triazole development by exposure of DAN to nitrite at pH 4 in the presence and in the absence of cafestol and kahweol up to 1 mM concentration indicated a detectable inhibitory effect of kahweol, although lesser than that of caffeic acid but little or no effect of cafestol (Figure 2). To assess in more detail the overall relevance of the furan diterpenes as nitrosation inhibitors, in further experiments, the reaction with acidic nitrite was performed on a ternary mixture containing cafestol, kawheol, and caffeic acid as the representative, most active coffee polyphenol, at 1 mM concentration each, under the conditions of the DAN assay. HPLC analysis after

30 min indicated a significant decay of caffeic acid, kawheol, and cafestol (92, 90, and 45%, in that order, Figure 3) and the formation of a main peak not attributable to caffeic acid decomposition (as assessed in a control mixture with caffeic acid alone under the conditions of the DAN assay). Overall, these data support the view that in the stomach, during digestion, nitrite ions from saliva and the diet may cause structural modifications of kahweol and, to a lesser extent, cafestol in coffee even in the presence of caffeic acid and related antinitrosating polyphenols, such as chlorogenic, ferulic, and p-coumaric acids (33, 34). On the basis of these results, in further experiments, the major reaction products formed by reaction of kahweol and cafestol with nitrite ions under acidic conditions were investigated. Acid-Promoted Reaction of Kahweol with Nitrite. The reaction of 1 mM kahweol with 1 mM nitrite in phosphate buffer at pH 3 resulted in complete substrate consumption after 2 h and the formation of a main product (Figure 4), which did not change over prolonged periods of time. In subsequent experiments, it was found that the same compound is formed with varying nitrite concentrations, at substrate/nitrite ratios ranging from 1:1 to 1:10. It remained the main product even under conditions of relatively high dilution, for example 100 µM. Extraction of a preparative scale reaction mixture with ethyl acetate followed by HPLC fractionation led to the isolation of the main product in pure form for structural characterization. The ESI+/MS spectrum displayed [M + H]+ and [M + Na]+ pseudomolecular ion peaks at m/z 331 and 353, respectively. These data were suggestive of a structure differing from the parent compound by an oxygen atom. The 1H NMR spectrum exhibited as a noticeable feature a doublet at δ 10.05 (J ) 5.0 Hz) coupled to a doublet at δ 5.87. Additional proton resonances were present at δ 7.44 and 5.95 (J ) 10.2 Hz). Inspection of the 13C NMR spectrum revealed four sp2 carbons and two carbonyl groups, one of which (δ 192.8) gave a cross-peak in the 1H, 13C HSQC spectrum with the downfield doublet at δ 10.05, suggesting a conjugated aldehyde function. Overall, these data denoted a profound alteration of the kahweol furan ring

1926

Chem. Res. Toxicol., Vol. 22, No. 12, 2009

De Lucia et al.

Chart 2

Figure 5. HPLC elution profiles of the reaction mixture of cafestol (1 mM) with nitrite ions (10 mM) at pH 3 (trace a, 0 min; trace b, 2 h). Elution conditions are described in the Experimental Procedures. Detection at 220 nm.

Table 2. 1H and 13C NMR Data of 1 (Acetone-d6) and 2 (CD3OD)a 1 1

H (mult, J)

1

c

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 C10-CH3

b

2 13

C

7.44 (d, 10.2)

163.2

5.95 (d, 10.2)

126.7

2.77 1.69 1.72 1.67

(m) (m) (m) (m)

1.51 (m) 1.78 1.92 1.54 1.78 2.08 1.76 1.88 1.50 1.63

(m) (m) (m) (m) (m) (m) (m) (m) (m)

3.55 (d, 11.2) 3.70 (d, 11.2) 5.87 (d, 5.0) 10.05 (d, 5.0) 1.20 (s)

190.5 156.9 54.2 22.7 40.2 45.8 49.3 48.3 19.7 26.3 45.8 38.2 53.3 81.1 66.4 132.3 192.8 19.0

H (mult, J)b

1

1.32 1.87 1.75 2.27 2.38 1.62 1.75 1.70

(m) (m) (m) (m) (m) (m) (m) (m)

1.35 (m) 1.68 1.82 1.61 1.77 2.12 1.78 2.05 1.52 1.70

(m) (m) (m) (m) (m) (m) (m) (m) (m)

3.67 (d, 11.2) 3.78 (d, 11.2) 5.63 (s) 0.88 (s)

13

C

36.1 33.4 88.7 166.0 47.6 22.2 41.1 45.6d 55.0 45.3d 20.2 26.9 46.3 38.6 53.8 82.8 66.8 115.4 169.0 14.8

a

Chemical shift values are given in ppm. b J expressed in Hz. Numbering of the parent compounds was retained as shown in Chart 1. d Interchangeable. c

that was consequent to a ring-opening and oxygen-insertion process. On the basis of extensive 2D NMR analysis, the compound was eventually formulated as 1 (Chart 2) (23% isolated yield). The lack of detectable cross peaks in the ROESY spectrum between the proton R to the aldehyde carbonyl and the methylene protons at C-6 allowed assignment of the E configuration to the exocyclic double bond. The complete spectral characterization of 1 is reported in Table 2. Careful analysis of the reaction mixture did not show other detectable reaction products, and no evidence was obtained for the formation of nitrogen-containing adducts even with higher nitrite concentrations, up to a 10-fold excess, the ring-opened aldehyde being invariably the main product. Product 1 was identified as the product eluting at 40 min in the HPLC trace of the reaction mixture of equimolar caffeic acid, kawheol, and cafestol with nitrite under the conditions of the DAN assay (Figure 3). Acid-Promoted Reaction of Cafestol with Nitrite. As previously noticed, cafestol was less reactive than kahweol

toward acidic nitrite. Accordingly, the reaction was investigated under a variety of conditions, and eventually a 10-fold excess nitrite was necessary to obtain sufficient substrate conversion and product formation. Two main products were detectable on HPLC (Figure 5), one of which proved to be inhomogeneous and unstable and was not investigated further. The other compound, less retained on the reverse phase column, was collected and was found to consist of a single species pure enough for spectral analysis. The ESI+/MS spectrum displayed [M + H]+ and [M + Na]+ pseudomolecular ion peaks at m/z 364 and 386, respectively, indicating the presence of a nitrogen-containing group derived evidently from nitrite incorporation. The 1H NMR spectrum exhibited a single resonance in the aromatic/olefin proton region (singlet at δ 5.63) that denoted a profound alteration of the furan ring. This conclusion was supported by the 13C NMR spectrum where a single CH resonance was apparent at δ 115.4 accompanied by a set of three quaternary carbon signals at δ 88.7, 166.0, and 169.0. These data, as supported by 2D correlation experiments, might be suggestive of a conjugated carbonyl flanked by a hemiketal/hemiaminal type quaternary carbon. Two alternative ring-closed structures might be envisaged viz. one featuring an oximino and hemiketal functionality and another showing a cyclic hydroxamic acid giving rise to a hemiaminal group. Careful analysis of the chemical shift pattern eventually led us to assign the product the 1-hydroxy-2-pyrrolinone structure 2 (Chart 2) (15% isolated yield). The complete spectral characterization of 2 is reported in Table 2. Mechanistic Issues. The higher reactivity of kahweol toward acidic nitrite was clearly due to the conjugated double bond, which not only enhanced the susceptibility of the furan ring to electrophilic attack but exerted also a role in diverting the reaction pathway from the expected nitrosation route, as seen in cafestol, to an alternate oxidative fission channel. To address in more detail the mechanism of the oxidative ring opening, in a separate set of experiments the reaction of kahweol with nitrite at pH 3 was carried out under a rigorously oxygen-free atmosphere. Analysis of the mixture after 2 h showed complete consumption of the substrate but no trace of the ring fission product 1, a finding that indicated a role of oxygen in the process. The importance of the acidic medium for product formation was supported by the very slow formation of 1 at pH 3 in air-equilibrated buffer even in the absence of added nitrite (20% yield after 2 h) and by other experiments in which kahweol was oxidized with CAN at pH 3 to give aldehyde 1 in good yield (about 30% formation yield as estimated by HPLC analysis). On the other hand, only small amounts of 1 were detected upon oxidation of kahweol with the Fenton reagent (Fe2+/EDTA/H2O2) at pH 7.4, an established model of biological nonenzymatic oxidations. These data overall indicated that both an acidic pH and an oxidant were required for the oxidative ring opening of kahweol and that a nitrous acid-derived oxidant

Reaction of Cafestol and Kahweol with Nitrous Acid

Chem. Res. Toxicol., Vol. 22, No. 12, 2009 1927

Scheme 1. Proposed Mechanism of Formation of 1 by Reaction of Kahweol at pH 3

Scheme 2. Proposed Mechanism of Formation of 2 by Reaction of Cafestol with Nitrite Ions at pH 3

was involved in the nitrite reaction. Comparison with cafestol, which was little or not affected at pH 3 in the absence of nitrite or in the presence of the Fenton reagent, confirmed the role of the conjugated double bond as a main determinant of the furan ring reactivity in kahweol. On this basis, a plausible mechanistic scheme accounting for the oxidative conversion of kahweol to 1 is proposed (Scheme 1). Initial fast protonation of the furan ring would give a cation intermediate, which, following addition of water, generates a hemiketal in equilibrium with ring-opened isomeric allylic alcohols through a highly conjugated enolic tautomer (only representative species are shown in the scheme). Aldehyde formation would then result from an oxidation step, for example, via nitrosonium attack to the OH group followed by a carbonylforming loss of nitroxyl (NO-) from the nitrite ester. Prevalence of protonation over nitrosation can be explained assuming that the conjugated double bond exerts a specific stabilizing effect on the rapidly generated protonated intermediate and allows the establishment of a series of equilibria that are driven by the irreversible oxidation step. Oxidative ring opening of 2,5- and 4,5-disubstituted furans to enediones has previously been reported to occur in the presence of peroxides and with different catalysts, under conditions that are relatively more forcing than those described herein for kawheol (35, 36). In the case of cafestol, on the other hand, the protonation equilibrium might be disfavored by the lack of efficient stabilization of the resulting cation and the ensuing intermediates by the conjugated double bond. Operation of relatively more forcing conditions is therefore necessary for the formation of a nitrosation-derived product (Scheme 2). Key steps of the proposed reaction pathway involve the addition of water to the

initially generated nitroso intermediate, the subsequent tautomerization to an oxime derivative, and the final equilibration to the 1-hydroxy-2-pyrrolinone product, via a ring-opened ketohydroxamic acid intermediate. The formation of a 1-hydroxy-2-pyrrolinone system by nitrosation of a furan derivative is not unprecedented, since it was recently observed in the pathway to a remarkable dimeric product obtained by reaction of menthofuran with acidic nitrite (24). Toxicological Relevance. The observed facile degradation of kahweol to an unsaturated aldehyde product under conditions resembling those occurring during digestion, especially if in the presence of elevated nitrite levels, raised an issue of whether such reaction could have toxicological implications associated to the chemical reactivity of the main product. To gain a preliminary insight into this issue, we briefly investigated the properties of the aldehyde 1 in model mutagenicity tests based on reaction with DNA bases in vitro. Under conditions commonly used for testing potential mutagen and DNA reactive compounds (27), with compound 1 at ca. 1:50 molar ratio with respect to 2′-deoxyguanosine 5′-monophosphate at pH 4.0, HPLC monitoring did not show appreciable consumption of 1 even after 24 h of reaction time. Likewise, no adduct formation could be observed when the reaction of 1 with 2′-deoxyguanosine was investigated under the conditions used for R,β unsaturated aldehydes (enals) derived from lipid peroxidation (28, 37). Aldehyde 1 also proved to be stable in the presence of acidic nitrite and exhibited no significant nitrite-scavenging capacity in the DAN assay at concentrations up to 1 mM. The same conclusions apply to compound 2 from cafestol. Thus, on the basis of these preliminary data, there seems to be no chemical ground to postulate a toxic effect of kahweol degradation by acidic nitrite. On the contrary, although the performance of kahweol in the DAN assay would support only a contributory role of this furan as an inhibitor of mutagenic and carcinogenic nitrosamine formation in the stomach, the observed degradation to species apparently devoid of adverse effects might be of relevance for the evaluation of the health implications of elevated intake of coffee furans. Indeed, the present results would suggest that during digestion, especially in the presence of nitrite, kahweol may be degraded, reducing its ingestion relative to that expected on the basis of the initial beverage composition and decreasing the actual cholesterol-elevating effects of the brew. This may be relevant to the current concern

1928

Chem. Res. Toxicol., Vol. 22, No. 12, 2009

about the possible negative impact of coffee furans on human health and the associated habit of filtering coffee brews to remove oily components putatively linked to heart disease and stroke. Acknowledgment. This work was carried out in part with the financial support of Regione Campania L5 (2006) to M.dI. Supporting Information Available: NMR instrumental details, 1H, 13C, COSY, TOCSY, ROESY, 1H, 13C HSQC, and 1 H, 13C HMBC NMR spectra of kahweol and compounds 1 and 2; HPLC elution profiles and MS spectra of cafestol and kahweol obtained from green coffe beans; HPLC elution profiles of the reaction mixture of coffee with nitrite ions at pH 3. This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Kamm, W., Dionisi, F., Fay, L.-B., Hischenhuber, C., Schmarr, H.G., and Engel, K.-H. (2002) Rapid and simultaneous analysis of 16O-methylcafestol and sterols as markers for assessment of green coffee bean authenticity by on-line LC-GC. J. Am. Oil Chem. Soc. 79, 1109– 1113. (2) Kurzrock, T., and Speer, K. (2001) Diterpenes and diterpene esters in coffee. Food ReV. Int. 17, 433–450. (3) Ricketts, M.-L., Boekschoten, M. V., Kreeft, A. J., Hooiveld, G. J. E. J., Moen, C. J. A., Muller, M., Frants, R. R., Kasanmoentalib, S., Post, S. M., Princen, H. M. G., Porter, J. G., Katan, M. B., Hofker, M. H., and Moore, D. D. (2007) The cholesterol-raising factor from coffee beans, cafestol, as an agonist ligand for the farnesoid and pregnane X receptors. Mol. Endocrinol. 21, 1603–1616. (4) Terpstra, A. H. M., Katan, M. B., Weusten-van der Wouw, M. P. M. E., de Roos, B., and Beynen, A. C. (2000) The hypercholesterolemic effect of cafestol in coffee oil in gerbils and rats. J. Nutr. Biochem. 11, 311– 317. (5) de Roos, B., Caslake, M. J., Stalenhoef, A. F., Bedford, D., Demacker, P. N., Katan, M. B., and Packard, C. J. (2001) The coffee diterpene cafestol increases plasma triacylglycerol by increasing the production rate of large VLDL apolipoprotein B in healthy normolipidemic subjects. Am. J. Clin. Nutr. 73, 45–52. (6) Cavin, C., Holzhaeuser, D., Scharf, G., Constable, A., Huber, W. W., and Schilter, B. (2002) Cafestol and kahweol, two coffee specific diterpenes with anticarcinogenic activity. Food Chem. Toxicol. 40, 1155–1163. (7) Tao, K.-S., Wang, W., Wang, L., Cao, D.-Y., Li, Y.-Q., Wu, S.-X., and Dou, K.-F. (2008) The multifaceted mechanisms for coffee’s antitumorigenic effect on liver. Med. Hypotheses 71, 730–736. (8) Cavin, C., Mace, K., Offord, E. A., and Schilter, B. (2001) Protective effects of coffee diterpenes against aflatoxin B1-induced genotoxicity: Mechanisms in rat and human cells. Food Chem. Toxicol. 39, 549– 556. (9) Cavin, C., Bezencon, C., Guignard, G., and Schilter, B. (2003) Coffee diterpenes prevent benzo[a]pyrene genotoxicity in rat and human culture systems. Biochem. Biophys. Res. Commun. 306, 488–495. (10) Majer, B. J., Hofer, E., Cavin, C., Lhoste, E., Uhl, M., Glatt, H. R., Meinl, W., and Knasmueller, S. (2005) Coffee diterpenes prevent the genotoxic effects of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) and N-nitrosodimethylamine in a human derived liver cell line (HepG2). Food Chem. Toxicol. 43, 433–441. (11) Kim, J. Y., Jung, K. S., and Jeong, H. G. (2004) Suppressive effects of the kahweol and cafestol on cyclooxygenase-2 expression in macrophages. FEBS Lett. 569, 321–326. (12) Kim, J. Y., Jung, K. S., Lee, K. J., Na, H. K., Chun, H.-K., Kho, Y.-H., and Jeong, H. J. (2004) The coffee diterpene kahweol suppress the inducible nitric oxide synthase expression in macrophages. Cancer Lett. 213, 147–154. (13) Kim, H. G., Kim, J. Y., Hwang, Y. P., Lee, K. J., Lee, K. Y., Kim, D. H., Kim, D. H., and Jeong, H. G. (2006) The coffee diterpene kahweol inhibits tumor necrosis factor R-induced expression of cell adhesion molecules in human endothelial cells. Toxicol. Appl. Pharmacol. 217, 332–341. (14) Lee, K. J., and Jeong, H. G. (2007) Protective effects of kahweol and cafestol against hydrogen peroxide-induced oxidative stress and DNA damage. Toxicol. Lett. 173, 80–87. (15) Bartsch, H., Ohshima, H., Pignatelli, B., and Calmels, S. (1992) Endogenously formed N-nitroso compounds and nitrosating agents in human cancer etiology. Pharmacogenetics 2, 272–277.

De Lucia et al. (16) Coss, A., Cantor, K. P., Reif, J. S., Lynch, C. F., and Ward, M. H. (2004) Pancreatic cancer and drinking water and dietary sources of nitrate and nitrite. Am. J. Epidemiol. 159, 693–701. (17) Lide, D. R., Ed. (1995) CRC Handbook of Chemistry and Physics, 75th ed., CRC Press, Boca Raton, FL. (18) Wink, D. A., Kasprzak, K. S., Maragos, C. M., Elespuru, R. K., Misra, M., Dunams, T. M., Cebula, T. A., Koch, W. H., Andrews, A. W., and Allen, J. S. (1991) DNA deaminating ability and genotoxicity of nitric oxide and its progenitors. Science 254, 1001–1003. (19) Lin, H., and Hollenberg, P. F. (2001) N-Nitrosodimethylaminemediated formation of oxidized and methylated DNA bases in a cytochrome P450 2E1 expressing cell line. Chem. Res. Toxicol. 14, 562–566. (20) Zenser, T. V., Lakshmi, V. M., Schut, H. A., Zhou, H. J., and Josephy, P. D. (2009) Activation of aminoimidazole carcinogens by nitrosation: Mutagenicity and nucleotide adducts. Mutat. Res. 673, 109–115. (21) O’Donnel, V. B., Eiserich, J. P., Chumley, P. H., Jablonsky, M. J., Krishna, N. R., Kirk, M., Barnes, S., Darley-Usmar, V. M., and Freeman, B. A. (1999) Nitration of unsaturated fatty acids by nitric oxide-derived reactive nitrogen species peroxynitrite, nitrous acid, nitrogen dioxide, and nitronium ion. Chem. Res. Toxicol. 12, 83–92. (22) Oldreive, C., and Rice-Evans, C. (2001) The mechanisms for nitration and nitrotyrosine formation in vitro and in vivo: Impact of diet. Free Radical Res. 35, 215–231. (23) Hegesh, E., and Shiloah, J. (1982) Blood nitrates and infantile methemoglobinemia. Clin. Chim. Acta 125, 107–115. (24) De Lucia, M., Mainieri, F., Verotta, L., Maffei, M., Panzella, L., Crescenzi, O., Napolitano, A., Barone, V., Appendino, G., and d’Ischia, M. (2007) Nitration Vs. nitrosation chemistry of menthofuran: Remarkable fragmentation and dimerization pathways and expeditious entry into dehydromenthofurolactone. J. Org. Chem. 72, 10123–10129. (25) Tricker, A. R., Haubner, R., Spiegelhalder, B., and Preussmann, R. (1988) The occurrence of tobacco-specific nitrosamines in oral tobacco products and their potential formation under simulated gastric conditions. Food Chem. Toxicol. 26, 861–865. (26) Kono, Y., Shibata, H., Kodama, Y., and Sawa, Y. (1995) The suppressing of the N-nitrosating reaction by chlorogenic acid. Biochem. J. 312, 947–953. (27) Lakshmi, V. M., Hsu, F. F., Schut, H. A., and Zenser, T. V. (2006) Stability and reactivity of 2-nitrosoamino-3,8-dimethylimidazo[4,5f]quinoxaline. Chem. Res. Toxicol. 19, 325–333. (28) Zhang, S., Villalta, P. W., Wang, M., and Hecht, S. S. (2007) Detection and quantitation of acrolein-derived 1,N2-propanodeoxyguanosine adducts in human lung by liquid chromatography-electrospray ionization-tandem mass spectrometry. Chem. Res. Toxicol. 20, 565–571. (29) Wettstein, A., Spillmann, M., and Meischer, K. (1945) The structure of cafestol. VI. HelV. Chim. Acta 28, 1004–1013. (30) Bertholet, R. Preparation of cafestol by selective hydrogenation of its mixtures with kahweol using a deactivated palladium catalyst. U.S. Patent 4692534, 1987. (31) Lam, L. K. T., Yee, C., Chung, A., and Wattenberg, L. W. (1985) Use of silver nitrate impregnated silica cartridges in the separation of kahweol and cafestol esters by preparative liquid chromatography. J. Chromatogr. 328, 422–424. (32) Urgert, R., Van der Weg, G., Kosmeijer-Schuil, T. G., van de Boverikamp, P., Hovenier, R., and Katan, M. B. (1995) Levels of the cholesterol-elevating diterpenes cafestol and kahweol in various coffee brews. J. Agric. Food Chem. 43, 2167–2172. (33) Takahama, U., Ryu, K., and Hirota, S. (2007) Chlorogenic acid in coffee can prevent the formation of dinitrogen trioxide by scavenging nitrogen dioxide generated in the human oral cavity. J. Agric. Food Chem. 55, 9251–9258. (34) d’Ischia, M., Panzella, L., Manini, P., and Napolitano, A. (2006) The chemical basis of the antinitrosating action of polyphenolic cancer chemopreventive agents. Curr. Med. Chem. 13, 3133–3144. (35) Massa, A., Acocella, M. R., De Rosa, M., Soriente, A., Villano, R., and Scettri, A. (2003) Mo(CO)6-Catalyzed oxidation of furan derivatives to E- and Z-enediones by cumyl hydroperoxide. Tetrahedron Lett. 44, 835–837. (36) Finlay, J., McKervey, M. A., and Nimal Gunaratne, H. Q. (1998) Oxidations catalysed by rhenium(V) oxo species 1. Conversion of furans to enediones using methyltrioxorhenium and urea hydrogen peroxide. Tetrahedron Lett. 39, 5651–5654. (37) Minko, I. G., Kozekov, I. D., Harris, M. T., Rizzo, C. J., Lloyd, R. S., and Stone, M. P. (2009) Chemistry and biology of DNA containing 1,N2-deoxyguanosine adducts of the R,β-unsaturated aldehydes acrolein, crotonaldehyde, and 4-hydroxynonenal. Chem. Res. Toxicol. 22, 759–778.

TX900224X