Impact of different pan-frying conditions on the formation of heterocyclic aromatic amines and sensory quality in fried bacon

Food Chemistry 168 (2015) 383–389

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

Impact of different pan-frying conditions on the formation of heterocyclic aromatic amines and sensory quality in fried bacon Monika Gibis ⇑, Miriam Kruwinnus, Jochen Weiss Department of Food Physics and Meat Science, Institute of Food Science and Biotechnology, University of Hohenheim, Garbenstrasse 21/25, 70599 Stuttgart, Germany

a r t i c l e

i n f o

Article history: Received 21 February 2014 Received in revised form 9 June 2014 Accepted 14 July 2014 Available online 22 July 2014 Chemical compounds studied in this article: IQ (PubChem CID: 53462) IQx (PubChem CID: 108354-47-8) MeIQ (PubChem CID: 62274) MeIQx (PubChem CID: 62275) 4,8-DiMeIQx (PubChem CID: 95896-78-9) 7,8-DiMeIQx (PubChem CID: 104855) PhIP (PubChem CID: 105650-23-5) Harman (PubChem CID: 486-84-0) Norharman (PubChem CID: 244-63-3)

a b s t r a c t Heterocyclic aromatic amines (HAAs) are formed in the crust of cooked meat products. Most HAAs are carcinogenic in long-term animal studies. Besides precursors in raw materials, important factors are temperature and heating time. Bacon slices were investigated for concentrations of HAAs after pan-frying under different monitored heating conditions. Two HAAs, MeIQx (2-amino-3,8-dimethylimidazo [4,5f]quinoxaline) (1.5–5.6 ng/g) and PhIP (2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine) (0.1– 2.6 ng/g), were found in pan-fried bacon slices. The bacon clearly contained higher concentrations of HAAs both with longer frying times and at temperatures of 200–220 °C rather than 150–170 °C, respectively. A similar continuous increase of the concentrations was observed for norharman (5.0–19.9 ng/g) and harman (0.3–1.7 ng/g). The sensory evaluation, using a hedonic test design for colour and flavour, of the pan-fried bacon slices resulted in a preferred frying time of 5 min at 150–170 °C. However, some testers clearly preferred crispy and darker bacon slices containing higher HAA concentrations. Ó 2014 Elsevier Ltd. All rights reserved.

Keywords: Heterocyclic aromatic amines MeIQx PhIP b-Carbolines Processed meat Bacon Pan-frying Creatin(in)e Sensory quality

1. Introduction Several epidemiological studies have shown that a relationship between diet and the risk of the incidence of cancer exists. Red and processed meat particularly focused on the risks of contracting colorectal cancer (WCRF/AICR, 2007). The second expert report by the World Cancer Research Fund (WCRF) and the American Institute for Cancer Research (AICR), in 2007, informed that there is a substantial amount of evidence that red meat and processed meat are convincing causes of colorectal cancer. Nevertheless, this report also noted that there was insufficient evidence to reach any consensus for nitrates, nitrites, heterocyclic aromatic amines (HAAs), nitrosamines, or polycyclic hydrocarbons as risk factors ⇑ Corresponding author. Tel.: +49 711 459 22293; fax: +49 711 459 24446. E-mail address: [email protected] (M. Gibis). http://dx.doi.org/10.1016/j.foodchem.2014.07.074 0308-8146/Ó 2014 Elsevier Ltd. All rights reserved.

for colorectal cancer. However, the International Agency of Research on Cancer (IARC) has given the recommendation to reduce daily intake of HAAs (IARC, 1993). HAAs are process contaminates which are generated in the Maillard reaction in the crust of meat products as a result of heat treatment. The amounts and ratios of precursors in the raw material, such as creatine, amino acids and reducing sugars, have an important influence on the formation of HAAs (Ahn & Gruen, 2005; Alaejos & Afonso, 2011; Jaegerstad, Skog, Arvidsson, & Solyakov, 1998; Skog, Johansson, & Jagerstad, 1998). The chemical interactions with antioxidants resulted in a reduction of HAA concentration in food (Damasius, Venskutonis, Ferracane, & Fogliano, 2011; Gibis & Weiss, 2012; Johansson & Jaegerstad, 1996; Murkovic, Steinberger, & Pfannhauser, 1998; Persson, Graziani, Ferracane, Fogliano, & Skog, 2003; Puangsombat, Jirapakkul, & Smith, 2011; Vitaglione & Fogliano, 2004) and model systems (Johansson & Jaegerstad,

384

M. Gibis et al. / Food Chemistry 168 (2015) 383–389

1996), and could inhibit the radical reactions in HAA formation which was showed in an electron paramagnetic resonance experiment (Kikugawa et al., 1999). In particular, the preparation method plays a key role in the formation of HAAs; frying and grilling in particular enhance the formation (Keating & Bogen, 2001; Knize, Salmon, Pais, & Felton, 1999). The most influential factors, of each preparation method, are the heating temperature and time (Ahn & Gruen, 2005; Arvidsson, Van Boekel, Skog, Solyakov, & Jagerstad, 1999; Skog et al., 1998). The formation of HAAs is expected to follow a first order reaction equation with the rate constant given by the Arrhenius equation (Arvidsson, Van Boekel, Skog, & Jagerstad, 1997; Arvidsson et al., 1999). Although meat and patties have been widely investigated with regard to their HAA content, only a small database exists about the common concentrations of HAAs for processed meat, such as bacon, sausages or ready-to-eat meat products (Puangsombat, Gadgil, Houser, Hunt, & Smith, 2011). In a recent study, salami and ham slices as pizza toppings were investigated, and sudden increases of HAA concentrations were found after exceeding specific temperatures and times (Gibis & Weiss, 2013). A recent review gives a detailed overview of this occurrence in meat products (Alaejos & Afonso, 2011). HAA levels for pan-frying of fried bacon were studied by some authors (Back, Lee, Shin, & Lee, 2009; Gross et al., 1993; Guy, Gremaud, Richoz, & Turesky, 2000; Johansson & Jaegerstad, 1994; Ni, McNaughton, LeMaster, Sinha, & Turesky, 2008; Puangsombat et al., 2011), but in most cases the bacon slices were bought from a grocery, so the raw material and thereby precursors may differ in the frying experiments. The objective of this study was the analysis of bacon slices concerning the concentration of HAAs after different monitored heating treatments. For this objective, the bacon was manufactured from belly without rind and gristle. As a working hypothesis, the concentration of HAAs may be affected by the pan-frying time and temperature. The bacon slices were manufactured according to the manufacturing process defined from one animal to exclude other factors of influence. The sensory preference of the cooked meat product is also a very important factor for the consumer intake of HAAs. In order to study this consumer preference, the sensory quality of the bacon slices according to colour and flavour, as well as liking, after pan-frying was determined using a hedonic test with a trained sensory panel.

was applied for data acquisition. The following items were used for HAA analysis: Ultra Turrax T-25 (IKA Labortechnik, Staufen, Germany), a centrifuge Biofuge 28 RS (Heraeus Sepatech, Osterode, Germany), an evaporator Barkley Model BB 74300 (Leopoldshöhe, Germany), a Model HP 8453 spectral photometer (Agilent Technologies, Waldbronn, Germany), solid phase extraction cartridges Bond ElutÒ PRS (propylsulfonic acid cation exchanger), Bond Elut C18, 100 mg, 500 mg, as well as (Varian, Palo Alto, CA) diatomaceous earth bulk sorbent Isolute HM-N, extraction blank cartridges IsoluteÒ from Separtis, (Grenzach-Wyhlen, Germany), a filter type 0967, 11 mm ID (Schleicher & Schuell, Dassel, Germany), and an analytical column TSK-gel ODS-80 250  4.6 mm, 5 lm (Tosoh Bioscience, Stuttgart, Germany), connected to a guard column Supelguard™ LC-18-DB (Supelco, Bad Homburg, Germany). The following equipment was used for the other determinations: extraction unit Büchi 810 (Büchi, Flawil, Switzerland) for fat, hydrolysing unit Büchi-425, and digestion unit Büchi-323 (Büchi, Flawil, Switzerland) for protein analysis, and Aqua Lab CX-2 (Decagon Devices Inc., Pullman, WA, USA) for measurements of water activity (aw). 2.3. Bacon preparation Belly without rind and gristle (approximately 2.5 kg initial weight) was used, and 40 g/kg curing salt (0.9% sodium nitrite, 99.1% sodium chloride), 5 g/kg bacon seasoning, 3 g/kg SchinkinÒ Cum Spezial (Reinert Gruppe Ingredients GmbH, Erftstadt, Germany) as curing aids, containing dextrose, monosodium glutamate, sodium ascorbate, sodium isoascorbate, maltodextrin, flavouring and 0.25 g/kg starter cultures (Bitec SM 96, Frutarom, Ditzingen, Germany). The preparation of the bacon was a dry-curing process for 4 days. After removal of the excess curing salt and a ripping time of 5 days (+2 °C, 75–80% rel. humidity), the bacon was dried for 4 h and smoked for 10 min in a universal smoking and heating chamber (Unigar 1800 BE Compact with software Ness Digitronic 4, Ness & Co. GmbH, Remshalden, Germany) using a smoking programme with 5 min cold smoke, and 5 min smoke settling. The pieces were dried again at 28 °C for 5 min and the process was repeated for up to 85.5% of the initial weight. Each piece of bacon was sliced into slices 1.5 mm thick. The slices were stored in modified atmosphere packaging (20% N2 and 80% CO2) at 5 °C until they were pan-fried.

2. Materials and methods 2.4. Heating conditions 2.1. Chemicals Norharman, harman and caffeine were purchased from SigmaAldrich (Taufkirchen, Germany). IQ, IQx, MeIQ, MeIQx, 4,8-DiMeIQx, 7,8-DiMeIQx, PhIP, Trp-P-1, Trp-P-2, Glu-P-1, Glu-P-2, AaC, and MeAaC were obtained from Toronto Research Chemicals (North York, Canada). Ultrapure water was produced on-site using an ultrapure water system (Herco, Freiberg, Germany). Aqueous ammonia (25%), acetonitrile (gradient grade), methanol (gradient grade), toluene, and ethyl acetate (gradient grade), were obtained from Carl Roth (Karlsruhe, Germany), and hydrochloric acid, ammonium acetate, sodium hydroxide, orthophosphoric acid, and triethylamine from VWR International (Darmstadt, Germany). All chemicals were analytical grade. 2.2. Apparatus and materials An HPLC system (Gynkotek, Germering, Germany), consisting of a M480 pump, Gina 50 autosampler, DG 1310 S degasser, RF 1002 fluorescence detector, UVD 320 diode array detector, and column oven (Thermo Technic Productions, Langenzersdorf, Österreich), was used. Gynkosoft chromatography data system Version 5.50

The bacon slices were pan-fried for 3, 4, 5, and 6 min at 150– 170 °C, as well as 2 and 3 min at 200–220 °C in a Teflon pan which was rubbed with sunflower oil before using. The pan was cleaned after each frying procedure and was rubbed again with sunflower oil. The temperature was monitored by using a data logger (Therm 3280-8M, Ahlborn, Holzkirchen, Germany). The bacon slices were continuously turned over after 1 min and 30 s, respectively, for last flip. The total pan-frying time for each side of each slice was equal. The heated bacon slices were then cooled and stored at 4 °C until use in subsequent analysis. 2.5. Determination of HAAs The method includes the determination of polar and apolar HAAs. The method of HPLC analysis (Gibis, 2007) used, with some modifications, was based on the method described by Gross and Grueter (1992). The bacon slices (n = 8) were cut with a blender and homogenised (approximately 16.5 g) with 90 ml sodium hydroxide solution (1 mol/l) using an Ultra-Turrax. Diatomaceous earth was added to each of the four equivalent amounts of homogenates and mixed. Two of the homogenates had been previously

385

M. Gibis et al. / Food Chemistry 168 (2015) 383–389

spiked with 100 ll of the standard mixture (IQ – 21.0, IQx – 21.6, MeIQ – 25.2, MeIQx – 14.7, 7,8-DiMeIQx – 11.8, 4,8-DiMeIQx – 12.0, Glu-P-2 – 8.4, Glu-P-1 – 8.2, norharman – 5.4, harman – 4.0, Trp-P-2 – 8.4, PhIP – 24.5, Trp-P-1 – 8.4, AaC – 6.2, and MeAaC – 6.1 ng/100 ll). HAAs were extracted from patties by multiple solid phase extraction. The HAAs were eluted with an ethyl acetate solution containing 5% toluene (100 ml) and adsorbed onto a coupled preconditioned cation exchanger PRS cartridge which had been previously preconditioned with a mixture of ethyl acetate and 5% of toluene. After the PRS cartridge had been dried and washed with hydrochloric acid (0.01 M HCl), the non-polar HAAs were eluted with a mixture of 15 ml solution of 0.01 M hydrochloric acid and methanol (2:3, v/v). In order to clean up the non-polar fraction, it was adsorbed onto a C18 cartridge (500 mg); after a cleaning step with ultrapure water and drying, the non-polar HAAs were eluted with a 1.2 ml mixture of methanol and ammonia (25%) (90:10, v/v) into a vial. After drying, the eluate was redissolved in 100 ll caffeine solutions (internal standard). The polar HAAs were eluted from the PRS cartridges with ammonia acetate (20 ml 0.5 mol/l, pH 8.5) and adsorbed onto a coupled C18 cartridge, which had been previously conditioned with methanol, for clean-up. After washing with ultrapure water (2 ml) and drying, the polar HAAs were eluted with a 0.8 ml ammonia (25%) methanol mixture (10:90, v/v), dried under nitrogen and redissolved in a 100 ll caffeine solution (internal standard). The flow rate was 1 ml/min. The mobile phase contained 10 mmol/l triethylamine with phosphoric acid adjusted to pH 3.2 as eluent A, 10 mmol/l triethylamine adjusted to pH 3.6 as eluent B, and acetonitrile as eluent C. The HAAs were separated with a gradient programme at 25 °C: 82–75% A, 10% B and 8–15% C from 0 to 10 min; 85–75% B and 15–25% C from 10 to 20 min; 0% A, 75–55% B and 25–45% C from 20 to 29 min; 0–82% A, 55–10% B, 45–8% C from 29 to 33 min; and 82–15% A, 10% B, 8–75% C from 33 to 35 min. The mobile phase with 75% C for 4 min was used for regeneration of the HPLC column. After regeneration, the HPLC system was set back to the original starting conditions, and equilibrated for 10 min. UV detection was carried out at 258 nm and a 3D field was used for spectra plots at 200–360 nm. The setting of the fluorescence detector (ex/em) was at 0 min (360/450 nm), 14 min (300/ 440 nm), 22 min (265/410 nm), 24 min (305/390 nm), 25.5 min (265/410 nm), and 28 min (335/410 nm). The peaks of the HAAs, norharman and Harman, were identified by comparing their retention times and UV spectra with those of the corresponding standards. Quantification was performed with an external calibration (norharman, harman) or single point standard addition (MeIQx, PhIP) (two samples and two with standard spiked samples). 2.6. Determination of principal components The creatine/creatinine content in fried beef patties was enzymatically analysed using an assay based on a method of Wahlefeld et al. (Roche diagnostics GmbH, Mannheim, Germany) (Wahlefeld, Holz, & Bergmeyer, 1974). The main composition of the prepared bacon was determined according to the instructions of the official collection of methods of food analysis using the §.64 LFGB methods (L 06.00-3 for dry matter/moisture, L 06.00-4 for minerals (ashes), L 06.00-6 for fat, L 06.00-7 for protein, L 06.00-8 for hydroxyproline (collagen), L 07.00-22 for glucose) (BVL, 2011). The content of sodium nitrite and nitrate (calculated as potassium nitrate) were determined as described by Roche Diagnostics GmbH (Roche, Mannheim, Germany) and the official German methods (L 08.0014) (BVL, 2011). 2.7. Determination of weight loss The three bacon slices were weighed together before and 1 h after pan-frying for the determination of weight loss. Four times

three slices were pan-fried at the same heating conditions and weighed. 2.8. Sensory evaluation Sensory panelists were trained for the sensory evaluation of bacon slices. The testers evaluated the heated slices after pan-frying. Sensory scores were rated using a continuous 10-cm scale for colour (0–10, very light to very dark, 5 = optimal) and for flavour (0–10, under-fried to over-fried (burnt), 5 = optimal). The panelists (n = 12) evaluated the samples by a mark on the scale. In addition, they had to mark the preferred sample. 2.9. Statistical analysis Means and standard deviations were calculated from these measurements using Excel (Microsoft, Redmond, VA, USA). Statistical analysis was carried out on the measurements of HAAs, water losses and colour values using the Statistic Analysis System (SAS 9.3, SAS Institute Inc., North Carolina, USA). In a first step, data were tested for normal distribution. In a second step, a variance analysis using Tukey’s test with the ‘‘Generalized Linear Model’’ (GLM) procedure was conducted to determine significant differences (a = 0.05) between sample batches. The levels of creatine and creatinine as function of heating time were analysed by linear regression analysis. 3. Results and discussion 3.1. Manufacturing process and composition of bacon before frying One possibility to generally improve the quality and shelf life of dry cured bacon and accelerate the manufacturing is to use starter cultures in the fermentation process. Fermentation may lead to a decreased growth of competitive spoilage microorganisms as well as food pathogens and improve food quality attributes, such as taste, aroma, texture, and colour. The use of starter cultures may lead to a noticeably shortened maturation time and the development of highly desirable food quality attributes. The main composition of the unheated bacon is shown in Table 1. The levels found in our analysis showed that the raw material of the bacon generally consisted of the principal compounds with little variation in their concentrations, including proteins, fats, minerals (ashes), moisture, sodium nitrite and potassium nitrate. These levels are typical concentrations and values for bacon. In the manufacturing process of the bacon used before pan-frying, the weight loss determined during the drying phase of bacon was found to be 5%, and the final weight after the smoking step was 85.5% of the initial weight. The bacon was manufactured using Table 1 Main composition of the bacon used before pan-frying (n = 3). Analysis

Mean

Standard deviation

Moisture (g/100 g) Protein (g/100 g) Fat (g/100 g) Minerals (ashes) (g/100 g) Sum of main compounds (g/100 g) Collagen (g/100 g) Nitrite (calculated as sodium nitrite) (mg/kg) Nitrate (calculated as potassium nitrate) (mg/ kg) Creatine (g/kg) Creatinine (g/kg) Glucose (g/kg) aw-Value

49.6 19.5 26.0 4.45 99.6 1.92 50 140

0.1 0.12 0.2 0.01 0.42 0.01 6.7 35

3.54 0.75 2.95 0.936

0.75 0.23 0.17 0.007

386

M. Gibis et al. / Food Chemistry 168 (2015) 383–389

strains of Staphylococcus and Micrococcus in a dry curing process. These facultative anaerobic strains have enzymatic activities of catalase, protease and lipase, and have the advantage of an improved reddening due to the activity of nitrate reductase, which can inhibit defects of colour and rancidity (Leroy, Verluyten, & De Vuyst, 2006). For this reason, we analysed the bacon used into its principle components, precursors of HAA formation and nitrite as well nitrate. Nitrite is known as an ingredient in an acidic milieu that can reduce the concentration of non-IQ-type HAAs in model systems (Tsuda et al., 1985). Nitrite and sodium ascorbate inhibited HAA formation in meat patties pan-fried at 225 °C for 10 min per side (Shin, 2005). The concentration of sodium ascorbate and isoascorbate (0.5 g/kg) was similar, and the sodium nitrite concentration was a third part in comparison to our study. Moreover, the addition of ascorbate has the advantage that the reddening was enhanced and the formation of nitrosamines was inhibited (Rywotycki, 2001). It causes fast changes of nitrites to nitrogen oxide and, at the same time, decreases the number of nitroso groups which may react with amines (Rywotycki, 2001).

3.3. Determination of creatine and creatinine in bacon before and after pan-frying

20

Creatine and the cyclised reaction product creatinine are important precursors of the formation of IQ-type HAAs and PhIP (Murkovic, 2004). The formation of imidazoquinolines and imidazoquinoxalines from products of the Maillard reaction (e.g. 2methyl-pyridine, 2,5-dimethyl-pyrazine) with acetaldehyde needs ˘ anadanovic´creatinine as a reaction partner (Milic´, Djilas, & C Brunet, 1993). The changes in the concentration of creatine and creatinine are shown in Fig. 2. Only a small amount of creatinine was found in bacon before pan-frying. The creatinine content of pan-fried bacon at a temperature of 150–170 °C increased linearly with the extended frying time from 0.07 to 0.14 g/100 g dry matter (Fig. 2B). The creatine content decreased with a higher temperature and increased frying times (Fig. 2A). The total creatine slightly decreased with increasing pan-frying time and temperature. The difference in the total creatine between unheated bacon and fried bacon slices increased to 0.1 g/100 g dry matter at the lower frying temperature with an extension of the frying time to 6 min, and to 0.2 g/100 g dry matter at the higher temperature and longest time. The formation and degradation reaction of creatinine may take place in a parallel way.

10

3.4. Determination of HAAs in bacon

3.2. Determination of weight loss after frying Similar weight losses in both temperature ranges were chosen for the standardization of the pan-frying conditions (Fig. 1). Weight loss of the bacon slices increased significantly with increasing frying time (p < 0.05). The weight loss increased at 150–170 °C from 39% to 53%, if the frying time was raised from 3 to 6 min. At 200–220 °C, an increase from 42% up to 52% occurred. Higher

60

d

d c

50

b

a

Weight loss (%)

heating times of 4 min resulted in the products all being burnt with weight losses of about 64%. Only bacon slices in the defined range of 39–53% weight loss were used for further tests, so that a sensory panel could additionally evaluate the products for flavour. The HAA contents were increased by increasing weight loss as a consequence of the longer frying times (Table 2). Similar results were found in a published study (Johansson & Jaegerstad, 1994): The weight loss during the frying of bacon was very high, being 50.3% and 71.4% for the moderate and well-done products, respectively. Puangsombat, Gadgil, Houser, Hunt, and Smith (2012) also reported that the cooking weight loss of fried bacon in this study was very high with a level of 71.9%, and that is one reason for the very high HAA levels in pan-fried bacon. The formation of PhIP also dramatically increased in cooking conditions that generated high weight loss (Murkovic, 2004).

b

40 30

70 °C 17 in 0 ,2 00 °C 3 m -2 in , 2 20° C 00 -2 20 °C 50 -

,1

m

2

70 °C

-1

m in

-1

50

,1

m in

5

6

-1 ,1

50 4

m

in

,1 in m 3

50

70 °C

0

Fig. 1. Weight loss (%) of the pan-fried bacon slices using different pan-frying conditions. Error bars indicated standard deviation (n = 4); means with different letters are significantly different (p < 0.05).

The analysis of HAAs in pan-fried bacon differed from the determination of fried patties described due to the content of moisture of approximately 50% before frying (Table 1) and the high weight loss of approximately 50% during pan-frying. For this reason, more sodium hydroxide per weighted sample was needed for the homogenisation step of the pan-fried bacon. The recovery rates of the HAAs were determined between 64.6% and 85.3% for polar HAAs and between 70.3% and 85.0% for the non-polar HAAs. The lowest recovery rate was found for PhIP (64.6%). Due to the

Table 2 Concentration of HAAs (mean ± standard deviation SD) in bacon after pan-frying using different temperatures and times. HAAs1

Frying conditions Temperature (°C)

Time (min)

MeIQx2 Mean ± SD(ng/g)

PhIP Mean ± SD (ng/g)

Norharman Mean ± SD(ng/g)

Harman Mean ± SD(ng/g)

150–170

3 4 5 6 2 3

1.9 ± 0.28a,c 1.5 ± 0.19a,c 1.8 ± 0.09a,c 4.9 ± 0.52b 2.4 ± 0.35c 5.6 ± 0.78b

0.2 ± 0.01a 0.1 ± 0.01a 0.1 ± 0.01a 1.1 ± 0.27b 1.3 ± 0.58b 2.6 ± 1.51c

5.0 ± 0.04a 8.4 ± 0.04b 9.9 ± 0.02b 14.1 ± 0.55c 13.7 ± 0.05c 19.9 ± 0.06d

0.4 ± 0.39a 0.7 ± 0.01a 0.7 ± 0.56a 1.7 ± 1.23b 1.3 ± 0.73b 1.6 ± 0.17b

200–220 1 2

Means with different letters are significantly different p < 0.05. No significant risk levels 0.41 lg per day OEHHA (2008).

Concentration of Creatin(in)e (g/100 g dm) co n m Un in h 4 , 15 ea t m in 0-1 ed 5 , 15 70 m ∞ in 0-1 C 6 , 15 70 m ∞ in 0-1 C 2 , 15 70 m ∞ in 0-1 C 3 , 20 70 m ∞ in 0-2 C ,2 2 0 00 ∞ -2 C 20 ∞ C

M. Gibis et al. / Food Chemistry 168 (2015) 383–389

1.0

A

Creatine Creatinine Total Creatine

0.8 0.6 0.4 0.2

Concentration of Creatin(in)e (g/100 g dm)

3

Ba

0.0

1.0

B

0.8 0.6 Creatine (R≤ = 0.83, p<0.001) Creatinine (R≤ = 0.88, p<0.001) Total Creatine (R≤ = 0.56, p<0.05)

0.4 0.2 0.0 0

1

2

3

4

5

6

7

Frying time at 150-170∞C (min) Fig. 2. Concentration of creatine, creatinine and total creatine (sum of creatine and creatinine calculated as creatine) (g/100 g dm) of the pan-fried bacon slices (A) using different temperature ranges for pan-frying; (error bars indicate the standard error of the mean (dm: dry matter)) (B) linear regression of creatine, creatinine and total creatine as a function of frying time at 150–170 °C (n = 10).

complex heated matrix, the polar fraction has to be clearly separated from the non-polar HAAs to avoid the co-elution of other non-polar compounds during the HPLC separation. MeIQx, PhIP, norharman, and harman were found in all fried bacon samples (Table 2). High contents of MeIQx were quantified after a pan-frying time of 6 min at 150–170 °C and after a frying time of 3 min at 200–220 °C. MeIQx concentrations of approximately 5 ng/g were found in both frying conditions. Both batches had a similar high weight loss of 52% and 53%, respectively; that is one reason for the high levels of MeIQx in pan-fried bacon. The lower temperature range of 150–170 °C used was the recommended temperature, and a temperature and time were suggested of 172 °C for 3 min on each side to minimise the nitrosamine level (Puangsombat et al., 2012). The concentration of PhIP was relatively low in all batches. However, the contents of PhIP increased with higher frying times (Table 2). Appreciable concentrations of PhIP were found by frying at a temperature range of 200–220 °C for 2 and 3 min. The concentration in the lower temperature range was half as much as in the longest heating time of 6 min. The concentrations of norharman and harman continuously enlarged both with the extension of frying time and increase of frying temperature (Table 2). A high concentration of norharman of 19.9 ng/g was quantified particularly at the higher frying

387

temperature. Norharman and harman have shown to possess comutagenic activities (Pfau & Skog, 2004). In addition to formation reactions, degradation of HAAs could also occur in parallel by frying at temperatures between 130 and 225 °C, subject to time (Arvidsson et al., 1997; Randel et al., 2007). Thereby, formation and degradation of HAAs correlated strongly to the type of frying fat and were also promoted by lipid oxidation products (Randel et al., 2007). The degradation stability is highest for the b-carbolines and the lowest for PhIP (Chiu & Chen, 1999). Comparable levels of 0.4–4.3 ng/g MeIQx and 0–4.8 ng/g PhIP were determined in an earlier HAA study, which was also performed after pan-frying of bacon (Sinha et al., 1998). A review carried out by Alaejos and Afonso (2011) summarised the HAA concentrations in different meats and processed meats. Both HAA compounds, MeIQx and PhIP, and the b-carbolines, norharman and harman, were found in comparable concentrations in bacon slices which were pan-fried, similar to our study, according to the frying temperature and time conditions. Additionally, 4,8-DiMeIQx was often quantified in very well-done products (Ni et al., 2008). Other authors also reported that the contents of HAAs in pan-fried bacon were very high and included 3.6 ng/g 4,8-DiMeIQx and 3.1 ng/g IQx (Puangsombat et al., 2012). The content of HAAs in pan-fried bacon was much higher than that of fully cooked or steamed bacon, which may be reduced due to the precooking process (Puangsombat et al., 2011). This industrial fully cooked bacon is cooked at a low temperature (162 °C) and with steaminduced high humidity either by using a continuous microwave oven or linear circulating oven (Puangsombat et al., 2012). In addition to the HAA concentrations in fried bacon, the mutagenicity was measured at about 2055 rev/g of bacon in the Ames test using the Salmonella typhimurium TA98 strain with metabolic activation by S-9 liver homogenate, obtained from Aroclor-1254 pretreated rats, whereas, the mutagenicity based upon the specific single mutagenic activity of each HAA compound and concentration in cooked bacon was calculated at a sum of 1700 rev/g (Gross et al., 1993). These approximations from data of the Ames test and chemical analyses indicated that PhIP, MeIQx and 4,8-DiMeIQx should be key contributors to the mutagenicity in cooked bacon (Gross et al., 1993). The content of HAAs in bacon varied significantly in most studies, which was mainly due to the factors of different raw material, amount and slice thickness of the bacon used for frying, or differences in the frying temperature. Gross et al. (1993) reported that IQ, MeIQ, Trp-P-1, Trp-P-2, Glu-P-1, Glu-P-2, and 7,8-DiMeIQx were not identified in any of the bacon samples analysed, and this finding was in agreement with our results. IQ, MeIQx and 4,8-DiMeIQx were found in similar amounts in moderately cooked bacon. In addition to the HAA levels in well-done fried bacon, the content of IQ was found to be very high, and the MeIQ and 4,8-DiMeIQx content was lower compared to a moderately cooked bacon (Johansson & Jaegerstad, 1994). The HAA content of the pan residue was also analysed with levels higher than in fried bacon slices; the ratios of HAAs in the fried bacon and the pan residues for the moderately and well-done bacon were found to be about 1:25 and 1:2, respectively (Johansson & Jaegerstad, 1994). 3.5. Sensory evaluation of fried bacon The fried bacon slices are shown in the images in Fig. 3. The panelists (n = 12) evaluated the different batches of hot bacon slices for colour and flavour directly after pan-frying. The sensory evaluation of the bacon slices resulted in an optimal evaluation regarding the bacon slices after pan-frying of 5 min for both colour and flavour (Fig. 4). The pan-fried slices which were heated for 5 min at 150–170 °C were the preferred samples of the testers

388

M. Gibis et al. / Food Chemistry 168 (2015) 383–389

3 min; 150−170°C

4 min; 150−170°C

of 5 min, it should be mentioned that some panelists often prefer more crispy bacon slices with a golden brown colour. One must consider that the content of MeIQx increased approximately 2.5 times for 1 min more pan-frying. The daily intake of these mutagenic substances may be increased. A guideline for the suitable pan-frying conditions should be clearly stated on the package of bacon slices so that the frying time chosen by the consumer is not too long and the temperature is not too high. 3.6. Key insights and potential explanation of results

5 min; 150−170°C

6 min; 150−170°C

2 min; 200−220°C

3 min; 200−220°C

Fig. 3. Photographs of the bacon slices after pan-frying using different temperatures and times.

10

c

c b

Flavour Colour

Sensory evaluation

8

b b

6

Optimal a,b a a a

a

a,b

a

In comparison to patties, the concentration of HAAs is very high in pan-fried bacon slices. One reason for this fact is the weight loss before pan-frying and increased weight loss during the pan-frying process. The moisture content is very low in the fried slices, which enhances the formation of HAAs. Furthermore, the thickness or high surface to weight ratio of the slices is also an important factor for the formation of HAAs. Normally, the slices only had a thickness of 1–2 mm. This enhances the weight loss during the frying process. Additionally, the fat is melted out and the meat part, containing typical precursors such as creatin(in)e, free amino acids and sugars, is enlarged, which may accelerate the HAA formation. Although ingredients with a known inhibitory effect on HAA formation, such as ascorbic acid and nitrite, occur in bacon, the concentrations of MeIQx and PhIP are very high in this type of meat product. Additionally, consumer preference plays a key role in the daily intake of HAAs. If individuals like more well-done meat products, such as crispy golden brown bacon stripes, the risk of a higher intake of these harmful substances is increased. For these reasons and to fully understand our results further, more detailed investigations, in particular the effect of typical ingredients in processed meat, are underway at the moment. 4. Conclusion In comparison to patties, higher concentrations of MeIQx and PhIP were found in pan-fried bacon due to the increased weight loss and decreased water activity during manufacturing and frying process enhancing the HAA formation. The sensory tests illustrated that the preference of colour and flavour plays a substantial role in consumers being exposed to these substances. If consumers prefer more well-done pan-fried bacon slices, the risk of a higher intake of HAAs will clearly increase.

4

Acknowledgements 2

3

m in ,1 50 4 m -1 in , 1 70 °C 50 5 m -1 in , 1 70 °C 50 6 m -1 in 7 ,1 0 °C 50 2 m -1 in , 2 70 °C 00 3 m -2 in , 2 20 °C 00 -2 20 °C

0

Fig. 4. Sensory evaluation of pan-fried bacon slices for the attributes colour (0–10, very light to very dark, 5 = optimal) and flavour (0–10, under-fried – very slight frying flavour to over-fried – burnt frying flavour, 5 = optimal/pleasant); means with different letters are significantly different (p < 0.05).

overall. The slices fried for 2 min at 200–220 °C produced similar results in the sensory evaluation. No significant differences with regard to colour or flavour were found between these two batches. Moreover, no significant difference was found for the batch which was pan-fried for 4 min at 150–170 °C. Although the sensory evaluation showed a preference for panfrying at a lower temperature of 150–170 °C and the frying time

The authors thank Silvia Lasta and Kurt Herrmann for their technical assistance. This project was supported by funds from the University of Hohenheim Experiment Station. Appendix . Abbreviations used: HAA, heterocyclic aromatic amines; SD, standard deviation; IQ, 2-amino-3-methylimidazo[4,5-f]quinoline (CAS No.: 76180-96-6); IQx, 2-amino-3-methylimidazo[4,5-f]quinoxaline (CAS No.: 108354-47-8); MeIQ, 2-amino-3,4-dimethylimidazo[4,5-f]quinoline (CAS No.: 77094-11-2); MeIQx, 2-amino-3, 8-dimethylimidazo[4,5-f]quinoxaline (CAS No.: 77500-04-0); 4,8DiMeIQx, 2-amino-3,4,8-trimethylimidazo[4,5-f]quinoxaline (CAS No.: 95896-78-9); 7,8-DiMeIQx, 2-amino-3,7,8-trimethylimidazo[4,5-f]quinoxaline (CAS No.: 92180-79-5); PhIP, 2-amino-1methyl-6-phenylimidazo[4,5-b]pyridine (CAS No.: 105650-23-5); Trp-P-1, 3-amino-1,4-dimethyl-5H-pyrido[4,3-b]indole (CAS No.: 62450-06-0); -monoacetate (CAS No.: 68808-54-8); Trp-P-2, 3amino-1-methyl-5H-pyrido[4,3-b]indole (CAS No.: 62450-07-1);

M. Gibis et al. / Food Chemistry 168 (2015) 383–389

Glu-P-1, 2-amino-6-methyldipyrido[1,2-a:30 ,20 -d]imidazole (CAS No.: 67730-11-4); Glu-P-2, 2-aminodipyrido [1,2-a:30 ,20 -d]imidazole (CAS No.: 67730-10-3); AC, 2-amino-9H-pyrido[2,3-b]indole (CAS No.: 26148-68-5); MeAC, 2-amino-3-methyl-9H-pyrido [2,3-b]indole (CAS No.: 68006-83-7); harman, 1-methyl-9H-pyrido[3,4-b]indole (CAS No.: 486-84-0); norharman, 9H-pyrido[3, 4-b]indole (CAS No.: 244-63-3).

References Ahn, J., & Gruen, I. U. (2005). Heterocyclic amines: 1. Kinetics of formation of polar and nonpolar heterocyclic amines as a function of time and temperature. Journal of Food Science, 70(2), C173–C179. Alaejos, M. S., & Afonso, A. M. (2011). Factors that affect the content of heterocyclic aromatic amines in foods. Comprehensive Reviews in Food Science and Food Safety, 10, 52–108. Arvidsson, P., Van Boekel, M. A. J. S., Skog, K., & Jagerstad, M. (1997). Kinetics of formation of polar heterocyclic amines in a meat model system. Journal of Food Science, 62(5), 911–916. Arvidsson, P., Van Boekel, M. A. J. S., Skog, K., Solyakov, A., & Jagerstad, M. (1999). Formation of heterocyclic amines in a meat juice model system. Journal of Food Science, 64(2), 216–221. Back, Y. M., Lee, J. H., Shin, H. S., & Lee, K. G. (2009). Analysis of heterocyclic amines and beta-carbolines by liquid chromatography–mass spectrometry in cooked meats commonly consumed in Korea. Food Additives and Contaminants, Part A, 26(3), 298–305. BVL (2011). Amtliche Sammlung von Untersuchungsverfahren nach § 64 LFGB, vorläufiges Tabakgesetz, §28b GenTG - Verfahren zur Probennahme und Untersuchung von Lebensmitteln, Bundesamt für Verbrauerschutz und Lebensmittelsicherheit (BVL). Berlin: Beuth Verlag. Chiu, C. P., & Chen, B. H. (1999). Stability of heterocyclic amines during heating. Food Chemistry, 68, 267–272. Damasius, J., Venskutonis, P. R., Ferracane, R., & Fogliano, V. (2011). Assessment of the influence of some spice extracts on the formation of heterocyclic amines in meat. Food Chemistry, 126(1), 149–156. Gibis, M. (2007). Effect of oil marinades with garlic, onion, and lemon juice on the formation of heterocyclic aromatic amines in fried beef patties. Journal of Agricultural and Food Chemistry, 55(25), 10240–10247. Gibis, M., & Weiss, J. (2012). Antioxidant capacity and inhibitory effect of grape seed and rosemary extract in marinades on the formation of heterocyclic amines in fried beef patties. Food Chemistry, 134, 766–774. Gibis, M., & Weiss, J. (2013). Formation of heterocyclic amines in salami and ham pizza toppings during baking of frozen pizza. Journal of Food Science, 78(6), C832–C838. Gross, G. A., & Grueter, A. (1992). Quantitation of mutagenic/carcinogenic heterocyclic aromatic amines in food products. Journal of Chromatography, 592(1–2), 271–278. Gross, G. A., Turesky, R. J., Fay, L. B., Stillwell, W. G., Skipper, P. L., & Tannenbaum, S. R. (1993). Heterocyclic aromatic amine formation in grilled bacon, beef and fish and in grill scrapings. Carcinogenesis, 14(11), 2313–2318. Guy, P. A., Gremaud, E., Richoz, J., & Turesky, R. J. (2000). Quantitative analysis of mutagenic heterocyclic aromatic amines in cooked meat using liquid chromatography–atmospheric pressure chemical ionization tandem mass spectrometry. Journal of Chromatography A, 883(1+2), 89–102. IARC (1993). Some naturally occurring substances: food items and constituents, heterocyclic aromatic amines and mycotoxins. IARC monographs on the evaluation of carcinogenic risk to humans (Vol. 56, pp. 163–242). Lyon: World Health Organization; International Agency for Research on Cancer. Jaegerstad, M., Skog, K., Arvidsson, P., & Solyakov, A. (1998). Chemistry, formation, and occurrence of genotoxic heterocyclic amines identified in model systems and cooked foods. Zeitschrift fuer Lebensmittel-Untersuchung und -Forschung A: European Food Research and Technology, 207(6), 419–427. Johansson, M. A. E., & Jaegerstad, M. (1994). Occurrence of mutagenic/carcinogenic heterocyclic amines in meat and fish products, including pan residues, prepared under domestic conditions. Carcinogenesis, 15(8), 1511–1518. Johansson, M. A. E., & Jaegerstad, M. (1996). Influence of pro- and antioxidants on the formation of mutagenic-carcinogenic heterocyclic amines in a model system. Food Chemistry, 56(1), 69–75. Keating, G. A., & Bogen, K. T. (2001). Methods for estimating heterocyclic amine concentrations in cooked meats in the US diet. Food and Chemical Toxicology, 39(1), 29–43.

389

Kikugawa, K., Kato, T., Hiramoto, K., Takada, C., Tanaka, M., Maeda, Y., et al. (1999). Participation of the pyrazine cation radical in the formation of mutagens in the reaction of glucose/glycine/creatinine. Mutation Research, Genetic Toxicology and Environmental Mutagenesis, 444(1), 133–144. Knize, M. G., Salmon, C. P., Pais, P., & Felton, J. S. (1999). Food heating and the formation of heterocyclic aromatic amine and polycyclic aromatic hydrocarbon mutagens/carcinogens. Advances in Experimental medicine and Biology, 459, 179–193. Leroy, F., Verluyten, J., & De Vuyst, L. (2006). Functional meat starter cultures for improved sausage fermentation. International Journal of Food Microbiology, 106(3), 270–285. ˘ anadanovic´-Brunet, J. M. (1993). Synthesis of some Milic´, B. i. L., Djilas, S. M., & C heterocyclic aminoimidazoazarenes. Food Chemistry, 46(3), 273–276. Murkovic, M. (2004). Formation of heterocyclic aromatic amines in model systems. Journal of Chromatography, B: Analytical Technologies in the Biomedical and Life Sciences, 802(1), 3–10. Murkovic, M., Steinberger, D., & Pfannhauser, W. (1998). Antioxidant spices reduce the formation of heterocyclic amines in fried meat. Zeitschrift fuer Lebensmittel Untersuchung und Forschung A/European Food Research and Technology, 207, 477–480. Ni, W., McNaughton, L., LeMaster, D. M., Sinha, R., & Turesky, R. J. (2008). Quantitation of 13 heterocyclic aromatic amines in cooked beef, pork, and chicken by liquid chromatography–electrospray ionization/tandem mass spectrometry. Journal of Agricultural and Food Chemistry, 56(1), 68–78. OEHHA. (2008). Proposition 65, status report, safe harbor levels: No significant risk levels for carcinogens and maximum allowable dose levels for chemicals causing reproductive toxicity. pp. 1–17: OEHHA (California Environmental Protection Agency: Office of Environmental Health Hazard Assessment), . Persson, E., Graziani, G., Ferracane, R., Fogliano, V., & Skog, K. (2003). Influence of antioxidants in virgin olive oil on the formation of heterocyclic amines in fried beefburgers. Food and Chemical Toxicology, 41(11), 1587–1597. Pfau, W., & Skog, K. (2004). Exposure to b-carbolines norharman and harman. Journal of Chromatography, B: Analytical Technologies in the Biomedical and Life Sciences, 802(1), 115–126. Puangsombat, K., Gadgil, P., Houser, T. A., Hunt, M. C., & Smith, J. S. (2011). Heterocyclic amine content in commercial ready to eat meat products. Meat Science, 88, 227–233. Puangsombat, K., Gadgil, P., Houser, T. A., Hunt, M. C., & Smith, J. S. (2012). Occurrence of heterocyclic amines in cooked meat products. Meat Science, 90, 739–746. Puangsombat, K., Jirapakkul, W., & Smith, J. S. (2011). Inhibitory activity of Asian spices on heterocyclic amines formation in cooked beef patties. Journal of Food Science, 76, T174–T180. Randel, G., Balzer, M., Grupe, S., Drusch, S., Kaina, B., Platt, K. L., et al. (2007). Degradation of heterocyclic aromatic amines in oil under storage and frying conditions and reduction of their mutagenic potential. Food and Chemical Toxicology, 45, 2245–2253. Rywotycki, R. (2001). Nitrosamine concentrations in beef ham: 1. Influence of smoking and diversified combinations of functional additives. Fleischwirtschaft, 81, 104–106. Shin, H. S. (2005). Influence of food ingredients on the formation of heterocyclic aromatic amine in cooked pork patties. Food Science and Biotechnology, 14(5), 572–575. Sinha, R., Knize, M. G., Salmon, C. P., Brown, E. D., Rhodes, D., Felton, J. S., et al. (1998). Heterocyclic amine content of pork products cooked by different methods and to varying degrees of doneness. Food and Chemical Toxicology, 36(4), 289–297. Skog, K. I., Johansson, M. A. E., & Jagerstad, M. I. (1998). Carcinogenic heterocyclic amines in model systems and cooked foods: a review on formation, occurrence and intake. Food and Chemical Toxicology, 36(9/10), 879–896. Tsuda, M., Negishi, C., Makino, R., Sato, S., Yamaizumi, Z., Hirayama, T., et al. (1985). Use of nitrite and hypochlorite treatments in determination of the contributions of IQ-type and non-IQ-type heterocyclic amines to the mutagenicities in crude pyrolyzed materials. Mutation Research, 147(6), 335–341. Vitaglione, P., & Fogliano, V. (2004). Use of antioxidants to minimize the human health risk associated to mutagenic/carcinogenic heterocyclic amines in food. Journal of Chromatography, B: Analytical Technologies in the Biomedical and Life Sciences, 802(1), 189–199. Wahlefeld, A. W., Holz, G., & Bergmeyer, H. U. (1974). Creatinine. Methoden der Enzymatischen Analyse, 2, 1834–1838. WCRF/AICR (2007). Food, Nutrition, Physical Activity, and the Prevention of Cancer: A Global Perspective. Washington DC: World Cancer Research Fund/American Institute for Cancer Research, AIRC.