Changes of Carotenoids in Atlantic Salmon by Heat Cooking and the Singlet Oxygen Quenching Activities of the Artificially Produced Carotenoids

Carotenoids are widely distributed in food such as vegetables, fruits, fish and crustacean animals, and are thought to play an important role in human health. Although the above materials are often heated for cooking, few studies have reported the change of dietary carotenoids by these processes. In this study, we analyzed the carotenoids in heat cooked (steamed, grilled, fried, and microwaved) Atlantic salmon fed mixtures of astaxanthin, adnirubin, and canthaxanthin, (6 : 3 : 1, all trans) (salmons ingested feed containing 80 mg Panaferd AX/kg) for two years, using a silica gel HPLC column, and compared with carotenoids contained in raw salmon for the first time. As a result, the cis-carotenods (9-cis astaxanthin, 13-cis astaxanthin, 13-cis canthaxanthin, 13-cis adonirubin) derived from salmon fed carotenoids were clearly increased in heat cooked salmon. The rates of cis-isomers/total (trans + cis-isomers) were microwave heating (21-32%), steaming and grilling (17-24%), and frying (14-21%), respectively.


Introduction
Carotenoids are C40 isoprenoid pigments (tetraterpenes) with long conjugated double bonds that possess colors ranging from yellow through to orange and red [1]. These pigments are widely distributed in food such as vegetables, fruits, fish and crustacean animals, and are thought to play important roles in human health, such as the prevention of cancer [2], metabolic syndrome [3], and eye disease [4], due to their antioxidative activity.
The Red color in wild Atlantic salmon (salmo salar) muscle derives from astaxanthin along with some minor carotenoids such as canthaxanthin, adonirbin, adonixanthin, zeaxanthin and antheraxanthin. These carotenoids originate from crustacean planktons, which are food for salmon, and salmon can store them in muscle and on the body surface. Also, salmon can reduce them on their body surface [5].
The structures of carotenoids change rapidly by oxidation or heat treatment. Since foods containing caroteonoids are often heat cooked for eating, studies on the change of carotenoids in vegetables (broccoli, spinach, green beans, cabbage, carrots, tomatoes and potatoes) [6,7] and fruits (orange, peach, mango and papaya) by heat cooking processes (microwaving, boiling and steaming) were reported previously [8][9][10][11]. On the other hand, only a few studies on carotenoid changes by heat treatment exist for seafood (edible fish, shell and crustaceans) have been reported [the change of the total carotneoid amount by heat treatment was reported for farmed (astaxanthin or canthaxanthin fed) rainbow trout in some studies [12,13].
In this study, we analyzed carotenoids in several heat cooked (steaming, grilling, frying and microwaving) Atlantic salmon, fed astaxanthin, canthaxanthin and adonirubin, using a silica gel HPLC column and compared to those contained in raw salmon. This is the first report comparing the changes of carotenoids by exhaustive heat treatments.

Preparation of atlantic salmon
Four male farmed atlantic salmon (1.335 kg, 1.367 kg, 1.478 kg, 1.644 kg) were fed Panaferd AX [astaxathin : adonirubin : canthaxanthin (6 : 3 : 1, all trans)] for 2 years (salmons ingested feed containing 80 mg Panaferd AX/kg) were used in this study. Two of them (1.367 kg and 1.478 kg) were used for carotenoids analysis in raw and cooked fillets, and the others (1.335 kg and 1.644 kg) were used to isolate carotenoids for 1 O 2 quenching activity. These were gifts from JX Nippon Oil and Energy Corporation (Japan, Tokyo). The muscle of each salmon were cut into 12 fillets (approximately 80 g/ fillet) and stored in a -80°C freezer.
water for 12 min, Group C: heated on a medium open grill for 15 min, Group D: fried in oil at 180-200°C for 2 min, and Group E: microwaved for 4 min at full power (500 W maximum output)]. The heat condition in each group was decided according to general cook books [14][15][16] to achieve over 75°C at the core.

HPLC analysis of carotenoids in cooked and raw salmon
The cooked fillets of each group (Group A-E) were cut into 15-20 cubes (2.0 cm×2.0 cm×2.0 cm, approximately), and extracted by stirring in 200 ml acetone for 1 hour at room temperature (×3 times). The extracts (600 ml) were added to hexane (600 ml) and H 2 O (600 ml), and partitioned in a 2 l separating funnel. The upper layer (red) was collected and dehydrated with sodium sulfate anhydrate, and the amount of total keto-carotenoids (astaxanthin, adonixanthin, adonirubin, and canthxanthin) in the upper layer was assessed using an Optical Density (OD) of 470 nm [extinction coefficient (absorbance of 1% concentration) of 2,100 was adopted for quantification] [17].

Statistical analysis
Data were analyzed using SPSS Statistics 22 software (IBM co.). Means of carotenoid value in each group (A-E) were compared by one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test (significance level was set at p<0.05). The percentage values shown in Table 1 were transformed using the arcsine transformation.

Spectroscopic analysis
1 H Nuclear Magnetic Resonance (NMR) spectra were measured with a Bruker AVANCE400. Chemical shifts were referenced to tetramethylsilane. Data processing was performed using Top Spin-NMR software (version 3.0) (Bruker BioSpin). High Resolution Electro Spray Ionization Mass Spectrometry (HRESI-MS) was recorded with a JEOL JMS-T100LP mass spectrometer. UV-VIS spectra were recorded with a Hitachi U-3200.
The identities and purities of the isolated compounds 1-8 were checked by HRESI-MS and 1 H NMR analysis. were added to micro glass vials (5.0 ml). The vials were tightly closed with a screw cap and a septum, and the mixtures were illuminated at 7,000 lux at 22°C for 3 hours in corrugated cardboard. Then, 100 µl of the reaction mixture was removed and diluted to 3.0 ml with ethanol, and absorbance at 235 nm was measured to estimate the formation of conjugates dienes [19]. The value in the absence of carotenoid was determined and 1 O 2 quenching activity was calculated relative to this reference value. Activity is indicated as the IC 50 (µM) representing the concentration at which 50% inhibition was observed. The IC 50 value was calculated by averaging the data from triplicate experiments.

Carotenoid composition in raw and cooked salmon
The results of carotenoids analysis (total amount and composition of each carotenoid) of raw and 4 cooked (steamed, grilled, fryied and microwaved) salmon are listed in Table 1. Although no significant differences were observed in the total carotenoids among Group A-E, cis-carotenoids (2, 4, 6+7) in salmon were significantly increased by heat cooking (Group B-E) (p<0.05). This observation indicates the conversion of the trans-carotenoid to the corresponding cis-isomers. The cis-isomer of 8 was inferred to have been produced, but it was not detected due to low productivity. Although the production of cis- isomers from trans-carotenoids on cooking of vegetables (tomato and sweet potato leaves) has been reported in some previous reports [20,21], this is the first report on the detailed changes of keto carotenoids (astaxanthin, adonirubin, adonixanthin and canthaxanthin) contained in fish and shellfish (Figure 1).
In this study, the heating time for group B-E was determined based on cookbooks [14][15][16] to examine the composition of carotenoids contained in cooked Atlantic salmon. The core temperature transition in salmon fillets by each heating method is shown in Figure 2, and the ratio of cis-isomers/total carotenoid (trans+cis) in each cooking treatment is shown as a bar graph in Figure 3. The ratio of each cis-isomer (2, 4, 6 and 7) was Group D (19-21%)<Group B and C (21-25%)<Group E (28-32%). The ratios of all cis-isomers in group E was significantly higher than some other groups (p<0.05) (Figure 3). Since the core temperatures by microwave cooking (88-94°C) were higher than those by steaming, open fire cooking, and frying (78-82°C) (Figure 2), these results are roughly explainable from their temperatures. Although the core temperatures of Group B, C and D were almost the same, the amount of cis-carotenoids was less in Group D than in Group B and C ( Figure 2) (There were no significant differences (p>0.05) between these groups). This observation may be explainable by the shorter heating time in Group D.
The ratios of cis-isomers in each heat cooking treatment were astaxanthin (14-21%) <adonirubin (19-27%) <canthaxanthin (21-32%) (Figure 3). Since it has been reported that the β-ionone ring of carotenoids in salmon muscle was combined with hydrophobic pockets of protein (such as actomyosin) and the presence of carbonyl (C=O) and hydroxyl (OH) groups enhance this combination [22], astaxanthin may be more stable than canthaxanthin in salmon.
In this study, the total carotenoid amounts (cis + trans) in Group B-E was the same as in Group A, indicating no chemical reactions except the isomerization from trans to cis-isomers. Since the decrease of total carotenoids (cis + tans) by excessive heating was observed in previous studies on vegetables [20,21], we suppose that the same decrease may also occur in Atlantic salmon if the heating time is increased.

Singlet oxygen quenching activity of carotenoids in raw and cooked salmon muscle
First, we examined the 1 O 2 quenching activity on crude carotenoid extracts (organic layer) of raw and cooked salmon. Their IC 50 values were calculated for 6.1 µM (Group A), 4.9 µM (Group B), 3.2 µM (Group C), 3.6 µM (Group D) and 2.6 µM (Group E). These observations indicated that cooking hardly influenced the 1 O 2 quenching activity of carotenoids when the total carotenoid amounts were not decreased.
Liu and Osawa reported that cis-astaxanthin shows more potent antioxidative activity than all-trans isomer in a DPPH radical scavenging assay and a lipid peroxidation assay using rat microsomes [24]. Clinton et al. described that more cis-carotenoids (such as lycopene) are taken up by the animal body than trans-isomer [25]. Further biological evaluations of the carotenoids in cooked Atlantic salmon considering these points are in progress.