Assessment of the N-nitrosopiperidine formation risk from piperine and piperidine contained in spices used as meat product additives
Eveline De Mey1*, Hannelore De Maere1,2, Lore Dewulf1, Hubert Paelinck1, Mieczysław Sajewicz3, Ilse Fraeye1, Teresa Kowalska3
1Research Group for Technology and Quality of Animal Products, KaHo Sint-Lieven, , Gebr. Desmetstraat 1, 9000 Ghent, Belgium, member of Leuven Food Science and Nutrition Research Centre (LFoRCe), Department M2S, Katholieke Universiteit Leuven, Kasteelpark Arenberg 20, B-3001 Leuven, Belgium
2ISA, Food Quality Laboratory, Boulevard Vauban 48, F-59046 LilleCedex, France
3Institute of Chemistry, University of Silesia, 9 Szkolna Street, 40006 Katowice, Poland
Corresponding Author: Eveline De Mey
E-mail address:
Tel: +32(0)9 265 86 10
Fax: +32(0)9 265 87 24
1
Abstract
The N-nitrosopiperidine formation in blends of spices andnitrite curing salt was investigated in relation with the piperine and piperidine contentsin spices.Firstly, two analytical methods were developed. Piperine was extracted with dichloromethane by means of accelerated solvent extraction, and determined by HPLC-DAD (λ=343 nm). A selective hydroextraction of piperidine usingASE and its quantification by HPLC-ELSD was applied. Both methods were sufficiently sensitive and accurate (limit of detection, limit of quantification and recovery: 0.28 µg, 0.84 µg, and 98.9±2.6% for piperine, and 5.76 µg, 17.45 µg, and95.9±2.9% for piperidine, respectively). Secondly, both compounds were quantified in commercial samples (black and white pepper, paprika, chili pepper, allspice and nutmeg). The maximum amount of piperine (21.12 mg g-1) was found in pepper, while the other spices contained only traces. Piperidine was detected mainly in the pepper samples, whereby the highest concentration was found in the white pepper extract (11.42 mg g-1). Thirdly, during the storage of spices blended with nitrite curing salt, the N-nitrosopiperidine content was determined,using agas chromatograph coupled with a thermal energy analyzer. Against our expectations, noN-nitrosopiperidine formation was observed in the curing mixture which contained white pepper extract. This result remains in contrast with the white pepper mixture, in which the N-nitrosopiperidine content significantly increased from not detected to 9.80±0.41 ngg-1 after the two monthsstorage period. In conclusion, highamounts of piperine or piperidine in spices do not systematically result in the formation ofN-nitrosopiperidine,when blended with nitrite curing salt.
Key words:piperyol-piperidine, piperidine, N-nitrosopiperidine, accelerated solvent extraction, spices, nitrite curing mixtures.
1
1. Introduction
Piperidine is a cyclic secondary amine, which can be considered as a parent molecular structure for many plant alkaloids, andpiperine (1-piperyol-piperidine) is the main pungent compound of pepper (Piper nigrum, Piperaceae) (Figure 1). White and black pepper, produced from the dried fruit without the pericarp and the dried whole, unripe berry, respectively[1], are used for many culinary purposes. The pungency, and thus the quality, of pepper (P. nigrum) are related to the amount of piperine [2]. However, this quality can be influenced by the hydrolysis of piperine, resulting in the cleavage of the piperidine ring [3].
In food (e.g, in spice premixes and the meat products), simultaneous presence of nitrite (as a curing agent) and pepper can give rise to the formation of the carcinogenic N-nitrosopiperidine (NPIP). This formation can occur either by the oxidative cleavage of the amide bond of piperine and a subsequent nitrosation of piperidine [4], or by a direct nitrosation of the already available piperidine [5,6]. In order to evaluate a risk of N-nitrosamine formation, it would be interesting to determine the piperine and piperidine contents in spices.
Normally, piperine and piperidine are extracted from ground pepper with use of organic solvents (e.g., methanol, ethanol, hexane, and dichloromethane). However, large quantities of these volatile solvents are consumed during the long extraction [7,8]. An optimized Soxhlet apparatus technique was proposed [9], but several alternative techniques have also been developed. For instance, high solubility of piperine in glacial acetic acid can accelerate the extraction [10], while the microwave-assisted extraction (MAE) enhances solvent penetration by breaking the cell structure of pepper[11]. Furthermore, ionic liquids can be used in ultrasonic assisted extraction (UAE)[12]. Although accelerated solvent extraction (ASE) is known to be a powerful technique to reduce the time and solvent consumption, reports concerning the extraction of piperidine alkaloids from spices are scarce. In fact, Rajopadhye et al. [13] described an ASE procedure for the extraction of piperine from the roots of Piper longum, while no reports were found about an application of this technique to isolation of piperidine from P. nigrum.
The most frequently used analytical techniques are thin-layer chromatography (TLC) and high performance liquid chromatography (HPLC). TLC is carried out onsilica plates, occasionally treated with silver nitrate, and usually with n-hexane or ethylacetate as eluents[14,15]. Although TLC is a relatively cheap method, it is less sensitive, as compared with HPLC. For the separation of the piperine isomers by HPLC, a C18 column is recommended[16]. Piperine analogues can directly be detected by the UVabsorption spectroscopy. However, piperidine is often derivatized with ninhydrin [17] or dansyl-chloride [18,19], due to the lack of a proper chromophoric group.
The aim of this study was threefold. Firstly, the methods for the determination of piperine and piperidine in spices were developed, in order to assess the quality of the commercial pepper samples. Secondly, the piperine and piperidine concentrations in a selection of culinary spices were measured. And finally, the N-nitrosopiperidine formation was evaluated during the storage of certain spices in combination with the nitrite curing salt. In that way,possible risk ofthe N-nitrosopiperidine formation was assessed, caused by piperine and piperidine containedat different concentration levelsin spices.
2. Experimental
2.1 Spices and nitrite curing mixtures
Different quality grades of the white and black pepper samples (P. nigrum) were selected. They also differed in particle diameter(e.g., thewhole peppercorns, the cracked pepper with the 12 mm particle diameter, and the pepper powderwith the ca. 0.3 mmaverage particle diameter were analyzed). The commercial powder was either produced by the fine grinding,or by evaporating ethanol from the spice extract. Besides of the pepper samples, some other spices were also purchased, namely paprika (Capsicum annuum) and chili pepper (Capsicum frutescens), both belonging to the plant family of the nightshades (Solanaceae), and also allspice (Pimenta dioica, Myrtaceae) and nutmeg (Myristica fragans, Myristicaceae). The spice samples were purchased either from the local supermarket(the black pepper powder and peppercorns came from Delhaize,Brussels, Belgium), or directly from the two suppliers.Nutmeg powder, white pepper powder and white pepper extract came from Raps (Kulmbach, Germany). The paprika, chili and allspice powder, and also the crackedwhite pepper and theMuntok white peppercornsoriginated from Rejo (Nazareth-Eke, Belgium).
In order to evaluate the risk of the N-nitrosopiperidine formation, meat curing mixtures were prepared. These premixes consisted of 3g of one of the above mentioned spices in 100 g nitrite curing salt (NaCl + 0.6% NaNO2, Rejo). These mixtures were vacuum packed and stored in the darkness at 25°C. For tracing the N-nitrosopiperidine formation, the mixtures were sampled (n=3) after one day, one week, and two months storage period.
2.2 Chemicals and standards
Hydrochloric acid (37% HCl), potassium hydroxide pellets (KOH), and antifoam silicone were all obtained fromVWR International (West Chester, PA, USA). Methanol and dichloromethane (DCM)were purchased from Merck (Darmstadt, Germany). All chemicals were of analytical grade. Water used in the experiment was de-ionized and double distilled in our laboratory by means of the Elix Advantage model Millipore system (manufactured in Molsheim, France).In order to prepare 3M KOH, a suitable amount of potassium hydroxide (KOH) was dissolved in de-ionized water. The 6M HCl solution was prepared by diluting a suitable volume of hydrochloric acid (HCl) with de-ionized water.
The piperine (≥97%), piperidine (≥99%),N-nitrosopiperidine (NPIP), and N-nitrosodipropylamine(NDPA) standardswere purchased from Sigma-Aldrich (St Louis, MO, U.S.A.). For the purpose of the HPLC analyses, piperine was dissolved in DCM at the concentration of 0.10 mg mL-1, and piperidine was dissolved in water at the concentration of 1.00 mg mL-1. NPIP was analyzed with a gas chromatograph coupled with a thermal energy analyzer (GC-TEA).For this purpose,the NPIP standard was dissolved and diluted in DCM to obtain a concentration of 0.25 µgmL-1. Quantification was carried out using NDPA as an internal standard (IS) at a concentration of 1 µgmL-1 DCM.
2.3 Accelerated solvent extraction
Prior to the HPLC analysis of piperine and piperidine, the ASE 200 model extractor (manufactured by Dionex, Sunnyvale, CA, U.S.A.) was used to carry out the accelerated solvent extraction on the spice samples. A0.5-g portion of the dry spice was carefully weighed and placed in the stainless steel cell and underwent three consecutive extraction runs. A pressure of 100 atm was applied. The remaining working parameters were, as follows: the solvent volume, 24 mL; static time, 15 min; number of extraction cycles within one extraction run, 2. Three 24-mL portions of extract obtained from a single spice sample were merged in a 100-mL calibrated volumetric flask and filled up to 100 mL with DCM or water, respectively. These solutions were used for the chromatographic quantification of piperine and piperidine in the investigated spices.
In order to optimize the extraction temperature for piperine with DCM as an extraction medium, the white pepper powder from Raps (containing a high level of piperine)was used as a model spiceat six different working temperatures (50, 60, 65, 70, 75 and 80°C).For the isolation of piperidine, the temperature of hydroextraction was optimized and this time, the white pepper extract from Raps (containing a high level of piperidine) was used as a model spice at six different working temperatures (40, 45, 50, 55, 60, and 65°C).
2.4 Recovery study
In order to evaluate theefficiency of the ASE procedure, a recovery study was performed. Therefore, the ground black pepper fruit was used as a model sample for the preparation of the blank plant matrix sample. In order to exhaustively extract piperine and piperidine from this sample, the aforementioned ASE procedure with DCMand H2O, respectively, as a solvent was repeated six consecutive times.
One series of the blank matrix samples was spiked with piperine (and extracted with DCM), and the other series of the blank matrix samples was spiked with piperidine (and extracted with H2O). The obtainedconcentrations per one gram of the blank plant matrix were specified in Table 1. From each dried and spiked matrix, four 1-gram portions were weighed out and each portion underwent a single ASE extraction run. Then the four extracts obtained in parallel from the four equally spiked matrix samples were condensed in the flow of nitrogen at 40°C to the volume of 1 mL each, with use of the TurboVap LV model evaporator (manufactured by Zymark, Hopkinton, MA, U.S.A.). These condensed solutions were then utilized for the chromatographic recovery studies and standard deviation was calculated from the recovery repetitions.
2.5 Determination of piperine and piperidine by HPLC
The HPLC analyses were carried out using a Varian model 920 liquid chromatograph (Varian, Harbor City, CA, USA) equipped with a Varian 900-LC model autosampler, a gradient pump, a Varian model 330 DAD detector, a Varian 380-LC model ELSD detector, and the Galaxie software for data acquisition and processing. The analyses were carried out in the isocratic mode, using a Pursuit 5 C18 (5 µm particle size) column (250 mm × 4.6 mm i.d.; Varian). As mobile phase, we employed methanol at a flow rate of 0.5 mL min-1 in the isocratic mode. The retention times (tR) of piperine and piperidine were equal to 5.9 and 10.9 min, respectively.
The presence of an aromatic chromophore in piperine offers a possibility of using a diode array detector (DAD). For the calibration curve purpose, the peak heights at the wavelength of 343 nm(i.e., at the absorption maximum of piperine)were employed. The calibration curve for piperine was obtained for the aliquots ranging from 0.20 to 1.50 µg piperine (in the intervals of 0.10 µg). The chromatographic peak heights were plotted against the microgram amounts of piperine injected on to the column. For each individual aliquot, six repetitions were performed (n=6). The limit of detection (LOD) and the limit of quantification (LOQ) were calculated using the formulas LOD = 3.3 × SD/a, and LOQ = 10 × SD/a, where SD is standard deviation of the peak height (n=6) taken as a measure of noise, and a is the slope of the corresponding calibration curve (y = ax + b).
The HPLC results valid for piperidine were derived from the ELSD detector. The calibration curve for piperidine was obtained for the aliquots ranging from 2 to 7 μg piperidine (in the intervals of 1 μg piperidine). This time, the chromatographic peak heights were plotted against the microgram amounts of piperidine. For each individual aliquot, six sampling repetitions were performed (n=6). The calibration curve and the LOD and LOQ values were determined, as described earlier.
2.6 Determination of NPIP by GC-TEA
In order to determine the contents ofN-nitrosopiperidine, the method according to Drabik-Markiewicz et al. [20] was used, with some modifications. The spice samples (in the 5-g portions) were spiked with 500 µL NDPA, and mixed with 200mL 3 M KOH and 4mL antifoam. The volatile N-nitrosamines were extracted from the spices by means of vacuum distillation from the alkaline solution. Subsequently, the distillates were extracted with DCM and washed with 6 M HCl. Finally, the extract was concentrated to 100µL in a Kuderna-Danish apparatus (Sigma Aldrich).For the detection and quantification, a GC-TEA (Thermo Electron Cooperation, Rodano, Italy) was used. The 5-µL aliquots of the concentrated extracts were injected on a packed column (10% Carbowax 20M+ 2% KOH on Chromosorb WAW, 80/100 mesh, 1.8m, 2mm i.d. Varian, Middelburg, The Netherlands) and a chromatographic separation was carried out with use of argon as a carrier gas (25 mL min-1). The injection port was set at 175°C and the oven temperature was increased from 110°C to 180°C at the 5°Cmin-1 rate. The temperature of the interface and the pyrolizer of TEA were set at 250°C and 500°C, respectively. The retention times (tR) of NDPA and NPIP were equal to 6.3 and 10.4 min, respectively. The calibration curves for NPIP were obtainedfrom 7 aliquots ranging from 0.625 to 12.5 ng with NDPA as an internal standard. The relative chromatographic peak areas (area NPIP/area NDPA) were plotted against the relative amounts of NPIP (µg NPIP/µg NDPA). For each individual aliquot, six repetitions were performed (n=6). The limit of detection (LOD) and the limit of quantification (LOQ)were 0.68 and 2.04 ng, respectively.
2.7 Statistical analyses
The results were expressed as the means ± standard deviation. The linear least-square regression was used to calculate the intercepts (a), slopes (b) and coefficients of determinations (r2) of the respective calibration curves (Microsoft Office Excel 2007, Microsoft Corporation, Redmond, WA). Significant differences among the extraction temperatureswere assessed with a one-way analysis of variance (ANOVA, PASW Statistics 19.0.0, SPSS Inc.). Piperine and piperidine contents of the spices and the accumulation of the NPIP content in the blends were also subjected to one-way ANOVAs. Tukey’s honestly significant difference criterion (p<0.05) was used to compare the means.
3. Results and discussion
3.1 Evaluation of the HPLC procedure
The chemical structure of the two investigated alkaloids proved a decisive factor for the choice of the most suitable detection and thus, thequantification mode also. The presence of an aromatic chromophore in piperine allowed the use of a diode array detector (DAD). As a result, good sensitivity was obtained and since the wavelength of 343 nm was proven tobe very selective, no other interfering peaks could be seen in the chromatograms (Figure 2).
The lack of a chromophore in the chemical structure of piperidine excludes a possibility of using DAD, and enforcesan employment of the universal ELSD detector, which is sensitive to the analyte’s molecular weight. In that way, an additional derivatization step, as applied by several authors [17-19] could be avoided. Although the LOD and LOQ values for piperidine are by one magnitude order higher than those valid for piperine, it can be concluded that all the results concerning method characteristics are satisfactory from the analytical point of view (see Table 2).
Most frequently, piperine and piperidine are extracted from the plant material by the Soxhlet extraction. In this study,ASE was applied as an alternative extraction technique, in order to reduce the analysis time and the solvent consumption. As it can be seen from Figure 3, the extraction yield of piperine (measured after a single extraction run) increased, when elevating the working temperature of the ASE extraction. At the temperatures higher than 70°C, no more significant improvement of the extraction yield could be observed,so it was decided to employ the temperature of 70°C for the DCM extraction of piperine.Due to an excellent solubility of piperidine and practical insolubility of piperine (and the other alkaloids) in water, a selective hydroextraction method for the isolation of piperidine was elaborated. Complete extraction of piperidine was observed over the whole range of the employed temperatures (results not shown). For practical reasons, the temperature of 50°C was chosen for the hydroextraction of piperidine (as the lowest working temperature of the ASE 200 apparatus is set on 40°C). Thus, the hydroextraction of piperidine could be done at a temperature considerably lower than that employed for the extraction of piperidine with DCM.
In order to evaluate theaccuracy and precision of both extraction procedures, the recovery studies were performed as well. For that purpose, a blank plant matrix was preparedfrom the black pepper powder by thoroughly extracting piperine with DCM and piperidine with water, using the ASE technique. After the six consecutive runs, neither of the two compounds could be chromatographically detected in the final extract. Hence, after drying the extracted plant material,it could be used as a blank plant matrix for the recovery study. The obtained resultsshowed an excellent performance of the ASE extraction system and practically full recovery of the two compounds of interest from the spiked plant matrix (Table 1). Moreover, the recovery precision was confirmed by a good repeatability of the recovery results (RSD of ca. 23%). This successful outcome allowed using the calibration curves derived for the piperine and piperidine standards from the HPLC quantification experiments (without a tedious spiking of the blank plant matrix, in order to prepare the plant-material-based calibration curves).