Food Technology and Biotechnology

Sample preparation and sous-vide cooking

Frozen fillets of Atlantic mackerel (Scomber scombrus) supplied from Pelagia A.S. (Selje, Norway) in January 2018, were used as materials in the study. The fish fillets were thawed at (0±1) °C overnight prior to experiment. Two fillets placed in each of the BST-090 type bags from Three Seal Bags series (Rolf Bayer Vacuumverpackung GmbH, Veitsbronn, Germany) with the following parameters: thickness 90 μm, gas permeability (O₂, N₂ and CO₂) up to 60 cm³/(m²·day·bar) and water vapour permeability up to 4 g/(m²·day). The vacuum bags with the fish were further heat-sealed using a vacuum sealing machine (SuperMax 3000 S; Webomatic, Bochum, Germany). Temperature data logger type SL52T (Signatrol, Tewkesbury, UK) was inserted into vacuum bags with the fish fillets to monitor fluctuations of temperature over the sous-vide cooking and chilling experiment. The antioxidant-treated and untreated fish fillets were sous-vide cooked by applying different time-temperature treatments and subjected to further chilled storage according to the planned experimental design displayed in Table 1. Two types of commercial antioxidants in liquid form were used: Fortium TR25 (rosemary extract) and Fortium RPT40 (a rosemary extract and ascorbyl palmitate) from Kemin Food Technologies, Herentals, Belgium. The antioxidants were manually distributed over the whole fillet surface by gentle rubbing, and left for 1 h to be absorbed by the fish flesh. The added amounts of antioxidants were the following: 0.001 mg/g of Fortium TR25 and 0.002 mg/g of Fortium RPT40. The fish samples were cooked in two water baths (Grant, Wilton, UK) set at 70 and 80 °C for 10 and 20 min. Immediately after the heat treatment, they were rapidly chilled with solid ice and put in chilled storage (in a cold room) at (4±1) °C for 1, 3, 9 and 15 days (Table 1). On each sampling day, the chilled mackerel samples were taken out of the vacuum packages and analysed. Control samples (raw mackerel fillets) were thawed at 0 °C overnight, vacuum-packed along with other experimental samples, and analysed after one day of chilled storage at (4±1) °C. Changes in cooking mass loss, protein solubility and oxidation, as well as texture parameters as affected by different cooking regimes and storage duration were studied. Analyses were performed in three and four replicates for each vacuum package containing two mackerel fillets.

Chemical and physicochemical assays

Experimental design and statistical analysis

A factorial design with four independent variables was employed in the study (Table 1). The independent variables were: cooking temperature (t(cooking)/°C), cooking time (τ(cooking)/min), use of antioxidants and duration of chilled storage (τ(storage)/day) as shown in Table 1, while the response variables were the main physicochemical parameters of mackerel samples. Experimental runs were randomized for each sampling day to reduce the effects of unexpected variability on the observed response. The experimental plan was used to identify the influence of sous-vide cooking regimes, use of antioxidants and duration of chilled storage on quality parameters of mackerel fillets.

The quality changes in sous-vide cooked mackerel were assessed through multiple regression analysis. The experimental data was analysed with SigmaPlot software, v. 14 (19) to derive statistically significant regression models from all possible linear and quadratic interactions between variables. The adequacy of the derived models was assessed statistically by R-value.

Statistical significance of the experimental data was verified by using Student’s t-test and analysis of variance (ANOVA). The coefficients of determination for all parameters were considered to have a good fit of the obtained regression models at the 95% confidence level.

Principal component analysis (PCA) was carried out to establish the variations and relationships among quality parameters of sous-vide cooked mackerel on the basis of experimental results. The dry matter, cook loss, protein oxidation and solubility, as well as flesh hardness data were entered into the PCA procedure using SigmaPlot software, v. 14 (19).

Dry matter and ash analysis

Mass fraction of dry matter in sous-vide cooked mackerel samples varied over a wide range of values (34.6 to 52.0%) during the storage (Table 2). On the first day of chilled storage the mass fraction of dry matter in all sous-vide cooked mackerel samples was significantly higher (p<0.05) than in the initial raw mackerel fillets used as control (38.6%, data not shown) samples. During storage, in the majority of mackerel samples the mass fraction of dry matter decreased. One possible explanation for this phenomenon is that cooking at 70 and 80 °C caused the rupture of fish muscle cells in the analysed mackerel fillets, leading to higher water release in the form of cook loss right after the treatment, thus increasing the mass fraction of dry matter in the samples. However, the released liquid was reabsorbed after the 3rd day of chilled storage by the thermally denatured unfolded and gelled proteins (20), thus decreasing the dry matter content of sous-vide cooked mackerel samples during storage. No significant variation was found in ash mass fraction in differently treated (sous-vide cooking and use of antioxidants) mackerel samples.

Cook and drip loss analysis

According to Table 2, cook loss was significantly (p<0.05) reduced in most of the sous-vide cooked mackerel fillets by the end of chilled storage compared to raw mackerel (control samples, 4.53%, data not shown). These data are in strong agreement with the results of a previous sous-vide cooking experiment performed on Atlantic mackerel (7). The tendency for reduction of cook loss in sous-vide treated mackerel samples during chilled storage may be explained by the reabsorption of liquid released during cooking by unfolded myofibrillar proteins and its distribution between the intra- and extracellular spaces (20). To identify the main factors influencing the variations in cook loss in the fish samples (Fig. 1a) compared to raw mackerel fillets, the following regression equation was derived in terms of actual values, on the basis of multivariate regression analysis:CL=0.865+0.057·t(cooking)+0.069·τ(cooking)–0.159·τ(storage) (R=0.821, R²=0.674, p<0.001) /1/ Fig. 1

3D surface plots displaying the influence of sous-vide regime parameters on quality characteristics of Atlantic mackerel on: a) cook loss, b) relative increase in total carbonyls in water- and c) salt-soluble proteins, and d) relative decrease in water-soluble proteins, where CL/% is cook loss, τ(cooking)/min is duration of sous-vide cooking, τ(storage)/day is duration of chilled storage, w(Fortium TR25 or RPT40)/% is the mass of added antioxidant per product mass, Δw(TC)/% is relative increase in total carbonyl content and Δw(WSP)/% is relative decrease in water-soluble proteins

The model clearly shows that the duration of chilled storage had the biggest contribution to the decrease in cook loss, thus leading to reabsorption of water by the unfolded denatured fish muscle proteins (20). Contrary to chilled storage time, temperature and duration of sous-vide cooking led to an increase in cook loss in mackerel samples, thus decreasing the product quality. However, according to regression coefficients of the model (Eq. 1), their influence was more than twice lower than of the chilled storage time. Therefore, we can conclude that the moisture lost in form of cook loss after sous-vide cooking of mackerel fillets was partially reabsorbed by the thermally denatured proteins during chilled storage.

Protein oxidation analysis

Oxidation (Fig. 1b and Fig. 1c) and solubility (Fig. 1d) of sarcoplasmic and myofibrillar proteins of Atlantic mackerel are affected by sous-vide cooking regimes and duration of chilled storage. During chilled storage, the carbonyl content significantly (p<0.05) increased both in the sarcoplasmic and myofibrillar proteins of sous-vide cooked mackerel samples (Fig. 2). This increase can be partially explained by the interaction of proteins with lipid oxidation products and various pro-oxidants (iron, myoglobin, etc.) liberated from the heat-disrupted cells during cooking (21). According to our hypothesis supported by earlier studies, sous-vide cooking broke down muscle structures, causing unsaturated fatty acids to react with haem pigments and other pro-oxidants, promoting autoxidation in the fish systems (21). The oxidized lipids came further in contact with proteins, leading to their carbonylation.

Fig. 2

Total carbonyl content in: a) and b) water-soluble (sarcoplasmic) and c) and d) salt-soluble (myofibrillar) proteins of Atlantic mackerel cooked at 70 °C (a and c) and 80 °C (b and d)

The initial carbonyl content in raw mackerel was 11.52 and 15.27 nmol/mg of sarcoplasmic and myofibrillar proteins, respectively, and it increased by 90% in sous-vide cooked fillets at the end of chilled storage compared to control samples. To explain which process parameters influenced this drastic increase in experimental mackerel samples, the following regression models (Fig. 1b and 1c) were derived in terms of actual values:Δw(C_WSP)=27.505–5.221·w(A)–44.225·w(B)+2.138·τ(storage) (R=0.879, R²=0.689, p<0.001) /2/ Δw(C_SSP)=–58.946+1.485·t(cooking)–82.038·w(B)+2.375·τ(storage) (R=0.757, R²=0.673, p<0.001) /3/where Δw(C_WSP/SSP) is the relative increase in total carbonyls in water- and salt-soluble proteins of experimental samples of sous-vide cooked mackerel compared to control samples (%), w(A) is the mass fraction of antioxidant Fortium TR25 (rosemary extract) (%), w(B) is the mass fraction of antioxidant Fortium RPT40 (ascorbyl palmitate) (%), t(cooking) is sous-vide cooking temperature (°C), and τ(storage) is the duration of chilled storage (day).

From the regression equations displayed above, we can conclude that the addition of rosemary extract and ascorbyl palmitate strongly protected fish tissue from oxidation, while duration of chilled storage was the main factor promoting protein carbonylation in sous-vide cooked mackerel samples. The protective effect of rosemary extract and ascorbyl palmitate was much stronger than the negative influence of chilled storage. Thus, the values of the regression coefficients for antioxidants Fortium TR25 and RPT40 were 2.44 and 20.68–34.54 times higher (Eqs. 2 and 3) than the regression coefficients for duration of chilled storage, respectively. Also, commercial antioxidant Fortium RPT40 added in the mass fraction of 0.2% retarded protein carbonylation in a more effective way than 0.4% Fortium TR25. Moreover, mass fraction of Fortium TR25 is not included in Eq. 3, and thus did not have any positive influence on the decrease of myofibrillar protein carbonylation. In addition, temperature of the sous-vide cooking has the second largest contribution to the increase of total carbonyl content in salt-soluble proteins after chilled storage (Eq. 3). Generally, myofibrillar proteins were characterized by a significantly higher carbonyl content than sarcoplasmic proteins in all cooking regimes (Fig. 1). Similar results were obtained for raw Atlantic mackerel fillets subjected to chilled and frozen storage (12). This tendency can be explained by a higher susceptibility of salt-soluble proteins to denature both during processing and storage (22).

Protein solubility analysis

There were no significant (p<0.05) changes in the solubility of water-soluble (sarcoplasmic) and salt-soluble (myofibrillar) proteins of mackerel fillets subjected to cooking at 70 and 80 °C during storage (Table 2). However, solubility of sarcoplasmic proteins significantly (p<0.05) decreased in mackerel samples cooked sous-vide at 80 °C in comparison with cooking at 70 °C. The significant decrease in protein solubility can be explained by protein aggregation as a result of cross-linkages between polypeptides and proteins, occurring due to protein oxidation promoted by cooking at higher temperatures (5). At the same time, no significant difference was found in the solubility of myofibrillar proteins in the differently treated fish samples. However, to reveal all factors contributing to the decrease in solubility of sarcoplasmic proteins of sous-vide cooked mackerel during chilled storage, the following regression equation (Fig. 1d) was derived in terms of actual values:Δw(WSP)=73.874+0.231·t(cooking)+0.084·τ(cooking) (R=0.712, R²=0.608, p<0.001) /4/where Δw(WSP) is relative decrease in water-soluble proteins (%), t(cooking) is sous-vide cooking temperature (°C), and τ(cooking) is duration of sous-vide cooking (min).

According to the regression equation 4, the second important factor leading to the decrease in the solubility of sarcoplasmic proteins after cooking temperature is duration of sous-vide treatment. However, its effect is 2.75 times less meaningful than the effect of cooking temperature (Eq. 4). Thus, according to the data displayed in Table 2, a 10 °C increase in cooking temperature reduced 1.33- to 1.72-fold the solubility of sarcoplasmic proteins in sous-vide cooked mackerel at the end of chilled storage. The regression model displayed above has not revealed any influence of antioxidants on the changes in protein solubility during the experiment.

Texture analysis

According to Table 2, there was a small but insignificant decrease in the hardness of mackerel fillets after sous-vide cooking compared to control samples ((8.0±1.0) N, data not shown). This tendency is in agreement with previous studies performed on sous-vide cooked Atlantic mackerel (7) and can be explained by heat-induced solubilisation of connective tissue in the temperature range of 50–70 °C leading to the flesh tenderization. However, the destruction of myofibrillar network is not severe before reaching a temperature of 70 °C resulting in the flesh toughening (7, 9). Therefore, mackerel samples subjected to sous-vide cooking at 70 °C had significantly lower values of hardness than the mackerel fillets cooked at 80 °C on the first day of experiment (Table 2). Moreover, since more proteins are denatured as the cooking temperature rises, their water-holding capacity decreases leading to an increase in cook loss and subsequent muscle toughening. Thus, cooking at 70 °C leads to softening of the fish flesh due to aggregation of sarcoplasmic proteins forming a gel, whereas heat treatment at higher temperatures (80 °C and higher) makes the fish flesh tougher due to viscous flow in the fluid-filled channels between the muscle fibres and fibre bundles increasing the elastic modulus (9). However, hardness of most samples of the studied sous-vide mackerel fillets was gradually increasing during chilled storage.

Variations among quality parameters revealed by PCA

In order to identify the main variations among quality parameters of sous-vide cooked mackerel samples, as well as the distribution of mackerel samples depending on sous-vide cooking regimes, the experimental data (dry matter, cook loss, protein oxidation and solubility, as well as flesh hardness) were subjected to PCA analysis. Thus, the PCA of the fish samples was performed on mean values of seven quality parameters of each fish group (sous-vide cooking and storage regimes) aiming to identify similarities and differences. Projection of the variables (quality parameters) on the factor planes revealed that the 1st and 2nd axes of the PCAs display 37.16 and 32.59% of the total variance (Fig. 3), respectively.

Fig. 3

PCA score plot (a) of mackerel samples cooked sous-vide at 70 °C (blue dots) and 80 °C (white dots); the numbers represent the sample numbers from the experimental plan, and (b) projection of quality parameters of the samples on the 1st and 2nd factor planes

The PCA score plot was composed to determine whether sous-vide treated mackerel samples could be clustered pointing to similarities in quality parameters under the same cooking and storage regimes. The resulting PC1/PC2 score plot (Fig. 3a) discriminates the samples according to treatment conditions. The PCA revealed a temperature-related clustering of sous-vide cooked mackerel samples. Thus, according to the PCA score plot, data on mackerel quality parameters can be clustered on the basis of cooking temperature, and major changes in the quality characteristics are due to heating regime.

On the PCA factor plane, PC1 best explained the variation in carbonyls in water- and salt-soluble proteins (Fig. 3b), while PC2 the variability of protein solubility of water- and salt-soluble proteins. This tendency confirms the fact that an increase in protein carbonylation contributes to a decrease in solubility of sarcoplasmic and myofibrillar proteins in sous-vide cooked mackerel samples. Besides variation in protein solubility, PC2 best explains the variability of dry matter, cook loss and hardness: these variables appeared to be influenced by each other. However, it is clearly notable that hardness of the fish flesh is influenced by both the protein carbonylation and variations in dry matter and cook loss. Thus, we assume that heat denaturation and oxidation of the fish muscle proteins resulted in their reduced water-holding capacity, leading to an increase in cook loss and subsequent muscle toughening. This assumption is in agreement with previous studies of quality changes in raw and sous-vide cooked Atlantic mackerel during chilled storage (7, 12).