1. Introduction
Lipid oxidation is one of the most important processes leading to deterioration in meat quality and is often crucial for food product shelf life [
1]. The process is induced by oxygen in the presence of initiators, such as high temperature, light and metal ions and is related to the unsaturation degree of the substrate [
2,
3]. Therefore, food products with high content of polyunsaturated fatty acids (PUFAs) are more exposed to lipid oxidation. For example, rabbit meat has many desirable nutritional features as a result of its lipid component, characterised by low levels of fat and cholesterol, a high level of unsaturated fatty acids and a good ratio of
n-6/
n-3 PUFAs with respect to other types of meat [
4,
5]. However, the higher amount of PUFAs leads to increasing oxidation with a consequent decrease in product stability during storage and cooking [
3,
6].
In addition to altering taste and nutritional quality, lipid oxidation involves the formation of reactive and toxic compounds, which may pose a risk to the consumer [
2,
6]. The major primary products, hydroperoxides, which are essentially odourless, are relatively unstable and decompose into a wide variety of secondary compounds among which aldehydes, in particular malondialdehyde (MDA), are considered to be the most important due to their great contribution to the development of rancid off-flavours and odours [
6,
7]. One of the major strategies for preventing lipid oxidation in food is the use of antioxidants, and a wide variety of compounds, from synthetic to natural, such as vitamin E (VE) and spices, has been studied [
8,
9]. However, there is still uncertainty regarding the effectiveness of many antioxidants and the optimal quantity to be used. For example, while the role of VE in stabilising meat colour and reducing lipid oxidative processes has been widely studied and assessed, including in rabbit [
4,
10,
11], the effects of selenium (Se) have been less studied and are still not conclusive. Selenium dietary supplementation has been more extensively studied in cattle, pig and poultry [
12,
13,
14], but only a few studies have considered rabbits, despite their meat susceptibility to lipid oxidation.
Recently, some authors have described a positive effect of Se supplementation on oxidative status and mineral profiles in rabbits [
15,
16,
17], and Minardi et al. [
5] found a protective effect against oxidative events in both the
Longissimus dorsi and
Biceps femoris muscles. Other authors have also found in pigs a synergistic effect between the administration of Se in the diet and the accumulation of VE in muscles [
18]. Additionally in pigs, the synergistic effect between Se and VE in combined dietary administration was also described, where an increase in lipase activity was observed with a consequent enhancement in post mortem lipolysis processes [
19]. However, to the best of the authors’ knowledge, the mechanism of action of Se has not yet been fully clarified, and the effects on the oxidative processes have, for the most part, been evaluated by quantifying thiobarbituric acid reactive substances (TBARS).
In terms of the specificity of the oxidative process, the choice of the best oxidation markers is very important for testing the antioxidant capacity and the extent of the lipid substrate oxidation [
2]. In the food industry, the most widely used methods measure changes in primary or in secondary products. Therefore, when oxidation occurs at a more accelerated rate, lipid oxidation is best assessed by measuring the secondary products, mainly aldehydes and ketones, due to their quick accumulation [
20].
TBARS analysis is one of the oldest and most commonly used methods of assessing lipid oxidation in food, and it is based on the spectrophotometric determination of MDA, a product of oxidation, directly in the food [
2]. The TBARS test is often used to evaluate oxidative status owing to its simple procedure and its high correlation with sensory scores [
21]; however, it has often been the object of criticism. In fact, MDA only forms from fatty acid chains containing at least three double bonds, such as linolenic acid. In this way, the decomposition products of the peroxidation of linoleic and oleic acid are excluded. Moreover, the TBARS test is not specific for MDA since thiobarbituric acid (TBA) can react with other aldehydes, browning reaction products, protein and sugar degradation products, amino acids and nucleic acids [
2]. The TBARS test can be used to assess the extent of oxidation rather than to quantify MDA [
1,
22].
In recent years, different methods of recovering secondary products have been proposed. Some of these products are considered to be highly specific markers of the lipid oxidation of a particular polyunsaturated fatty acid family. For instance, propanal is regarded as the main oxidation marker of
n-3 fatty acids, while hexanal and pentanal are reputed to be good oxidation markers of
n-6 fatty acids. In particular, hexanal, as its formation is usually greater than that of other secondary products, is the most used marker for lipid oxidation [
6]. In several studies on food, solid-phase microextraction coupled with gas chromatography (SPME/GC) has been used to detect lipid oxidation products such as hexanal [
1,
22,
23,
24]. Although the use of this technique is increasing due to its sensitivity and simplicity, it is not always reported as the most suitable technique for detecting the extent of lipid oxidation, and it is still controversial as a replacement for the TBARS test. Some studies have found a perfect correlation between the SPME/GC technique and that of TBARS [
22], while in other cases, TBARS proved to be more sensitive and accurate oxidation markers [
1].
In this work, the effects of two antioxidants, VE and EconomasE
TM (EcoE), on lipid oxidation are considered by comparing two different analytical methods. Hence, the effect of EcoE, a patented premixture of nutritional additives consisting mainly of Se, was compared with that of VE in rabbit hamburgers during refrigerated storage. The study of antioxidant supplementation appears to be of great interest, in particular in meat preparations because they might be more easily exposed to oxidative processes as well as bacteria contamination, and cold storage is not always sufficient to contain these problems. The choice of rabbit meat was related to its valuable nutritional characteristics that, due to its greater richness in PUFAs, make it more susceptible to oxidative processes; moreover, rabbit meat has been studied much less than other meats despite its nutritional value and its increased use in recent years [
25].
In particular, the principal aim of this study was to evaluate the effects and the specificity of VE or Se dietary supplementation on the lipid stability of rabbit hamburger meat during refrigerated storage, measuring the secondary products of oxidation through TBARS spectrophotometric analysis and the SPME/GC determination of the hexanal content.
In parallel, the possible relationship between the data related to the peroxidation products and those related to meat colour and pH, quality parameters often related to oxidation processes, was considered. The sensory discriminability among cooked rabbit hamburgers obtained from the different dietary groups was also assessed using triangle tests.
2. Materials and Methods
2.1. Animals and Diets
The Scientific Ethics Committee on Animal Experimentation approved the experimental protocol (University of Bologna, Prot. ID 1/72/2012).
For the analyses, 270 commercial New Zealand white male rabbits (
Oryctolagus cuniculis), provided by the Rabbit Genetic Centre of the Martini Group (Budrio di Longiano, FC, Italy), were selected and randomly divided into 5 dietary groups of 54 animals each; these were subdivided into 3 groups with ad libitum access to water and feed, as reported by Albonetti et al. [
26] and Minardi et al. [
5].
Both the starter basal diet (SBD) and the finisher basal diet (FBD) used in the first (5–8 weeks) and second (9–12 weeks) periods, respectively, were formulated to meet the nutrient requirements of the animals during the experimental period and were supplemented with 50 mg/kg of VE (Sigma-Aldrich, St. Louis, MO, USA) and 0.22 mg/kg of sodium selenite (Na
2SeO
3) corresponding to a Se amount of 0.099 mg/kg of feed. Ingredients and compositions of the two basal diets are shown in
Table 1 and detailed in Minardi et al. [
5].
The basal diets of two groups were supplemented with 100 or 200 mg/kg of dl-α-tocopheryl acetate (Sigma-Aldrich, St. Louis, MO, USA) (VE 100 and VE 200 groups, respectively). The diets of another two groups were supplemented with 100 or 200 mg/kg of EcoE (Alltech Nicholasville, KY, USA) (EcoE 100 and EcoE 200 groups, respectively), as suggested by the producer. The remaining group was fed a normal diet and used as a control (CTRL). The organic Se in the EcoE premixture corresponded to a Se amount of 1500 mg/kg of feed; therefore, the 100 and 200 mg/kg added to the EcoE100 and EcoE200 diets contained 0.15 mg and 0.30 mg/kg of Se, respectively. In the EcoE premixture, the amount (50,000 mg/kg) of vitamin C (VC) corresponded to 5 and 10 mg/kg of feed in the EcoE100 and EcoE200 diets, respectively. The small amount of VC present in EcoE, lower than the minimum requirements for rabbits [
21], suggests excluding that any effects found with the administration of EcoE could be attributed to VC.
The contents of α-tocopheryl acetate and Se in feed are reported in
Table 2 as previously determined by Minardi et al. [
5].
At 12 weeks of age, the rabbits were transported to the slaughterhouse (Ma.Ge.Ma abattoir, Savignano sul Rubicone, FC, Italy), weighed and electrically stunned (70 V DC, 50 Hz, 5 s). After complete bleeding, skin and viscera were removed, and the hot carcasses were weighed and then cooled at 4 °C for 24 h in a refrigerated cell (Costan Daily TN SP60/4, Costan S.p.A., Belluno, Italy). Ten carcasses were randomly selected from each dietary group for analysis. Carcass hygiene was tested, as reported in detail in Albonetti et al. [
26], and the meat was used to prepare hamburgers.
2.2. Rabbit Hamburgers
Carcasses from each dietary group were boned, and the meat was minced twice using a refrigerated mincer (TC 32 Frozen, Sirman Spa, Padova, Italy) to generate a single batch of minced meat from which all the necessary hamburgers were produced, without any seasoning or additives in order to avoid interference with the tested antioxidants.
The hamburgers (100 ± 3 g; 3 for each group and for each sampling time) were immediately prepared, using a conventional burger maker having an average dimension of 10 cm in diameter and 1 cm in thickness, and were wrapped with plastic food film. A part of the minced meat, after being placed in a plastic bag and subjected to vacuum treatment, was stored at −20 °C and was used for sensory analysis as described in the specific section.
The hamburgers were stored at 0–2 °C in a refrigerator (Quartet 200, Costan, Belluno, Italy) which was closed with a lid every night to reproduce retail storage conditions. Immediately after production (time 0, t0) and after 1, 2 and 4 days (t1, t2 and t4, respectively), the hamburgers were analysed for pH and colour or were stored in liquid nitrogen if designated for lipid oxidation analyses. Preliminary tests had shown that this storage method effectively protects against oxidative processes.
The contents of α-tocopherol and Se in minced meat, previously determined by Minardi et al. [
5], are reported in
Table 2. The Se content in meat burgers of EcoE200 groups was 20.3 µg/100 g, well below the minimum daily requirement for humans (55 µg/100 g) [
27].
2.3. Oxidation Products Analysis
The occurrence of secondary oxidation products in hamburgers was checked using two different approaches, which complemented each other, i.e., the analysis of TBARS and the determination of the hexanal.
2.3.1. TBARS Analysis
The TBARS were determined using a colorimetric assay [
28] as equivalents of MDA, one of the low-molecular-weight products formed by lipid peroxidation [
29], using a UV-Vis Spectrophotometer Perkin Elmer Lambda 45.
The samples (2 g) were mixed with 20 mL of ultrapure water and homogenized using an Ultraturrax (IKA) for 30 s at 13,500 rpm. Aqueous trichloroacetic acid (25% w/v) was added to a 5% final concentration (v/v), and the samples, after stirring at 4 °C for 15 min, were centrifuged for 15 min at 11,500 rpm. An aliquot (2.1 mL) was pipetted from the supernatant into a screw-capped tube, and 0.9 mL of 0.6% w/w 2-TBA water solution was added. The mixture was incubated at 70 °C for 30 min and then cooled under tap water. The TBARS levels were estimated at 535 nm using malonaldehyde bis(dimethyl acetal) (1,1,3,3,tetrametoxypropan) as standard. The concentration of MDA in the samples was calculated on the basis of the slope and intercept data of the computed least-squares fit of the calibration curve.
2.3.2. Hexanal Analyses
The analyses to determine hexanal content were carried out on homogenised samples (1.0 g) put into 10 mL glass cylindrical auto-SPME vials (Supelco
®) to which a NaCl saturated solution was added to reach 5 mL [
30]. The vials, immediately sealed with an assembled aluminium screw cap with a hole and a polytetrafluoroethylene/silicone septum, were stirred for 1 min at room temperature, conditioned at 40 °C for 30 min and then placed in an autosampler at 50 °C for 5 min for absorption of the headspace volatiles using SPME. A mixed SPME fibre 50/30 µm divinylbenzene-carboxen-polidimethylsiloxane (DVB/CAR/PDMS, Supelco
®) was used to extract the headspace volatiles. The target analyte on the loaded fibre was thermally desorbed for 5 min at 250 °C directly in the GC injector port, used in splitless mode. Both the absorption and the desorption phases were controlled by the Varian CP-8200 autosampler. The GC analyses were carried out on a Varian 3380 GLC equipped with a fused silica capillary column Equity-5 Supelco (30 m × 0.25 mm i.d. and film thickness 0.25 µm) and a flame ionisation detector (FID). The chromatographic conditions were: H
2 as a carrier gas (20 mL/min); temperature program: 35 °C was held for 5 min, 8 °C min
−1 to 75 °C, 40 °C min
−1 to 200 °C and final isotherm at 200 °C for 5 min; the FID was held at 250 °C.
The data were processed using a Varian Star Chromatography Workstation. The identification of hexanal levels was carried out using internal standards. An external calibration procedure was used to quantify the extracted amount of target analyte from the SPME fibres, with a calibration range from 50 to 1000 ng. Linear concentration–response relationships were always obtained with high confidence (r2 > 0.95).
2.4. pH and Colour Evaluation
The ultimate pH (pHu) was measured using a pH-meter (HI92240, Hanna Instruments, Padova, Italy) with a penetration electrode (Double-Pore cod. n° 32384003, Hamilton). The pH probe was calibrated using two buffers (pH 4.01 and 7.01, HI5400-12 and HI5700-12, respectively, Hanna Instruments, Villafranca Padovana, Italy), and each measurement was performed three times.
The colour parameters (L*, a*, b*) were evaluated using a tristimulus analyser equipped with an 8 mm diameter measuring area (Minolta Chroma-Meter CR-200, Minolta Inc., Osaka, Japan) according to the CIE LAB approach [
31], selecting D65 as illuminant and a 0° viewing angle, as mentioned by Honikel [
32].
2.5. Sensory Evaluation
Triangle tests [
33] were carried out to determine whether plain consumers could detect any difference between the CTRL and VE 200 or CTRL and EcoE 200 dietary groups. The highest concentration of antioxidant was selected because it was thought that any effect on the sensory response would be more discernible. Moreover, in this way, at each detection time, it was possible to halve the number of consumers required to implement a “proper” triangle test, i.e., using a brand-new set of consumers with no previous experience with rabbit meat supplemented with antioxidants. In detail, on six distinct occasions (at 15, 30, 60, 120 and 240 days from the date of pouch preparation), six different groups of 18–20 consumers, as detailed in
Section 3.3, all regular consumers of rabbit meat and all volunteers, convened in three separate shifts in a rather large room. That room was both olfactorily and visually neutral and temporarily arranged with 6–7 individual booths kept in dim light. Each panel was composed of students and staff at the Department of Veterinary Medical Sciences - DIMEVET (overall gender ratio, male proportion = 0.474; age ranging from 24 to 67 years). Each consumer received a triangle test [
32] arranged on a separate tray, prepared with meat patties presenting either the CTRL vs. VE200, or the CTRL vs. EcoE200 dietary groups to be compared, as detailed below.
Ground meat from each dietary group had been prepared as 30 g patties, each of which was put into an uncovered glass baby food jar and microwaved at 750 W (cooking time = 120 + 120 + 60 s) in a 2450 MHz, 1000 W variable power oven Mod. MT 243/486, (Whirlpool Europe, S.r.l., Comerio, Italy). Each jar had been covered with its own lid to keep its contents as warm as possible and then randomly assigned a three-digit number. Each triangle test consisted of three jars (the contents of two of which were identical while the third was different) placed in a line on the same tray. The panellists were asked to identify the odd sample, having been previously instructed to cleanse their palate between samples with the soft part of sliced unsalted Tuscan bread and still water at room temperature. In addition, as suggested by the relevant International Organization for Standardization (ISO) norm [
33], the consumers were asked to briefly comment as to why their choice of the odd sample had been made.
2.6. Statistics
The data were reported as mean values ± standard deviation (SD) of at least three determinations. To compare the differences between the dietary experimental groups and between storage times, two-way analysis of variance (ANOVA) was carried out on all data. The Tukey test was used to determine if the differences in the mean values were statistically significant (p ≤ 0.05). Before the statistical analyses, the data had been checked for normal distribution and variance homogeneity. All the statistical analyses were carried out using SigmaStat® release 2.0 (SPSS Inc., Chicago, IL, USA).
As for sensory evaluation [
33], a significance level of
p ≤ 0.05 was selected to determine whether the number of correct answers was sufficient as given, each time, by a panel of 18–20 plain consumers.