Effects of Dietary Inclusion of Canthaxanthin- and α-Tocopherol-Loaded Liposomes on Growth and Muscle Pigmentation of Rainbow Trout (Oncorhynchus mykiss)

Institute of Natural Products Chemistry, Vietnam Academy of Science and Technology (VAST), 18 Hoang Quoc Viet Cau Giay, Hanoi, Vietnam Graduate University of Science and Technology, Vietnam Academy of Science and Technology (VAST), 18 Hoang Quoc Viet Cau Giay, Hanoi, Vietnam Research Institute for Aquaculture No. 1, Bac Ninh Province, Vietnam Hong Duc University, ,anh Hoa City, ,anh Hoa Province, Vietnam Faculty of Environmental and Food Engineering, Nguyen Tat ,anh University, Ho Chi Minh City, Vietnam NTT Hi-Tech Institute, Nguyen Tat ,anh University, Ho Chi Minh City, Vietnam Institute of Environmental Technology, Vietnam Academy of Science and Technology (VAST), 18 Hoang Quoc Viet Cau Giay, Hanoi, Vietnam Institute of Chemistry, Vietnam Academy of Science and Technology (VAST), 18 Hoang Quoc Viet Cau Giay, Hanoi, Vietnam


Introduction
In the market of farmed aquatics, reddish color is often perceived as indication of high quality fish flesh and is associated with consumer acceptance [1]. However, pigments responsible for the red coloration of fish flesh, comprising mostly carotenoids such as canthaxanthin and astaxanthin, can only be naturally synthesized in bacteria, algae, yeasts, molds, and some higher plants with de novo synthesis capacity [2,3]. erefore, supplementation of carotenoids in feeds has been commonly exercised in commercial farming of salmonids, crustaceans, and other aquatic organisms to give desirable coloration to the resulting products [4][5][6]. Carotenoids are a pigment class that is responsible for a wide range of colors in plant and animal species and figure prominently among other pigments as an important antioxidant primarily derivable from marine species [7]. It has been shown that dietary supplementation of canthaxanthin, a common carotenoid, in aquaculture salmon and trout could color the fillet with red-pink color [8,9] and that pigmentation intensity of rainbow trout flesh was positively correlated with added canthaxanthin in fish feed [10].
Carotenoid feeding in aquaculture is associated with several issues. First, production of carotenoids largely depends on chemical processes and thus presents certain disadvantages including strict requirements on process control and generation of undesirable by-products that might be allergic to some parts of consumers [11]. In addition, high-dose consumption of canthaxanthin has been reported to result in health issues, most notably the development of retinal crystals in human eye [12] due to canthaxanthin-induced damage in lipid membranes of macula blood vessels [13]. As a result, canthaxanthin addition in commercial fish feed is widely regulated, and the allowable dose per kilogram of salmonid feed is 25 mg according to the regulation of European Union [14]. e second obstacle is that carotenoids such as astaxanthin and canthaxanthin, due to their unsaturated chemical structures, have low aqueous solubility and are sensitive to physicochemical conditions such as heat, oxygen, and light, resulting in low bioavailability of the substances when being used as nutraceutical ingredients [15,16]. Indeed, Torrissen et al. [17] showed strong, inverse relationship between canthaxanthin dose in fish feed and apparent digestibility in rainbow trout, while increased dietary lipid levels increased its apparent digestibility [17]. erefore, further measures in aiding the distribution of drugs into target sites and prevention of impacts of negative environmental factors on the carotenoids are essential.
In this study, we approached the aforementioned concerns by first formulating liposomes loaded with active substances derived from natural sources, followed by evaluation of the effect of dietary supplementation of liposomal canthaxanthin on growth, fillet color, muscle deposition, and nutrient composition of fish flesh of cultured rainbow trout (Oncorhynchus mykiss). e use of liposome has been regarded as an efficient strategy for delivery of drugs with low bioavailability since it allows better oral adsorption of lipid insoluble polar compounds, in turn resulting in improved drug efficacy. Canthaxanthin was extracted from bacterium Paracoccus carotinifaciens, an emerging natural source that is known for accumulation of canthaxanthin at high yield, at around 0.4% [18]. In liposome formation, α-tocopherol was coloaded along with canthaxanthin to maintain the quality and color of the fish flesh [19][20][21][22], and soybean lecithins were used as phospholipids owing to their continuous availability and advantages in terms of cost, emulsifying behavior, and sensorial characteristics [23,24]. e inclusion of α-tocopherol is justified by its role in filling the gaps left by the imperfect packing of phospholipids in liposomal formation and in maintaining fish flesh quality, including the color of salmon, by affecting the oxidative stability of lipids [25]. Current findings are expected to justify the use of liposome containing natural-derived colorants in intensive aquaculture practices and greatly enhance the value of natural canthaxanthin given that consumers are increasingly aware of negative health effects that synthetic colorants may bring.

Materials. Bacterium
Paracoccus carotinifaciens VTP 20181 strain was isolated from a salt field land in Diem Dien town, ai Binh province, Vietnam, and fermented for 64 hours. Methanol (MeOH), acetone, n-hexane, dichloromethane, tetrahydrofuran (THF), 2-propanol, and acetonitrile solvents were obtained from Sigma Aldrich (St. Louis, MO, USA). Canthaxanthin standard was purchased from TRC Canada. A commercial preparation of soybean lecithin phospholipids was used to synthesize liposomes.

Extraction of Canthaxanthin. Biomass of the bacterium
Paracoccus carotinifaciens VTP20181 obtained after fermentation was provided by the Food Industry Research Institute (Hanoi, Vietnam). Biomass was refrigerated at -40°C prior to extraction process. Wet biomass was collected by centrifugation and washed with water, followed by drying.
en dried, canthaxanthin-rich biomass was extracted with solvent (ethanol 96% + 0.5% glycerol monostearate) under ultrasound assistance. Afterwards, the solvent in the mixture was removed by rotary evaporation to afford the canthaxanthin extract. e extract is crystallized with urea to remove saturated fatty acids. Subsequently, column chromatography was used to purify the extract of canthaxanthin. Finally, emulsification was carried out from the liquid, which was then dried to collect canthaxanthin powder [26]. e final yield was approximately 1.5 mg of canthaxanthin, accounting for 0.15 % of dry biomass weight. e compounds were elucidated using spectroscopic methods (IR, 2D-NMR and MS).

Liposome Synthesis and Characterization
2.3.1. Synthesis of Liposome. Canthaxanthin-and α-tocopherol-loaded liposomes were prepared by thin-film evaporation method as described previously [27]. Briefly, a mixture of canthaxanthin and α-tocopherol at the fixed mass ratio of 1 : 9 (mixture A) was dissolved in 2 mL of dichloromethane together with soybean lecithin. e initial concentrations (IC � m canthaxanthin /m lipids , % wt/wt) of canthaxanthin were selected at 0.1%, 0.5%, and 1.0%, respectively. After dissolution, the thin membrane was obtained by removing the organic solvent in the water bath at 30°C under flask pressure of 250 mbar and rotational speed of level 2. Distilled water (4 mL) was added to peel off the membrane, forming canthaxanthin-loaded liposomes. In order to get a fixed size, liposome was passed through a 100 nm polycarbonate membrane in a mini extruder 50 times. e final sample was sealed in a new eppendorf and kept in the refrigerator (about 4°C in the dark). All other samples (blank liposomes, canthaxanthin-loaded liposomes, and α-tocopherol-loaded liposomes) were prepared with the same procedure and respective ratios of materials. e structure of canthaxanthin-and α-tocopherol-loaded liposomes is shown in Figure 1.

Size Measurement.
e z-average, mean size, and polydispersity index (PDI) of the liposome samples were monitored by using a SZ-100 nanoparticle size and zeta potential analyzer (Horiba, Japan) with a He/Ne laser (λ � 633 nm) and scattering angle 90°. e obtained results are the average of at least three measurements.

Canthaxanthin Loading.
e encapsulation efficiency (EE%) and drug loading content (DL%) were determined by UV-Vis spectrophotometer. Unencapsulated canthaxanthin was removed by dialysis (14000 Da, 10 cm × 10 cm). e suspension was dialyzed for 24 hours against 1 L of distilled water. e distilled water was changed every 2 hours from the start 4 times.
e liposomal system was dissolved in 3.5 mL of hexane and sonicated in a bath for 2 min to destroy the liposomal suspensions. After that, 100 µL of liposomes mixture was combined with 2900 µL of hexane and vortexed for 30 seconds. e free amount of canthaxanthin was quantified spectrophotometrically by Hitachi U-2900 spectrophotometer (Hitachi, Tokyo, Japan) at 474 nm with hexane as a blank. e EE % and DL % were determined, respectively, by the following equations: EE% � the amount of canthaxanthin in the liposomes the total amount of canthanxanthin × 100 , DL% � mass of canthaxanthin in the liposomes mass of canthaxanthin/liposomes × 100. (1)

In Vitro Drug Release.
First, 1 mL of canthaxanthinand α-tocopherol-loaded liposomes was placed into dialysis bags (molecular weight cutoff � 14000 Da), followed by incubation in 50 mL of phosphate buffered solution (PBS) (pH � 7.4) containing Tween 80 (0.5% wt) at 37°C with gentle shaking (100 rpm). After certain time period, 1 mL of sample was taken from the bag, and an identical volume of fresh buffer was added. Withdrawn media were extracted and analyzed by using UV-Vis spectrophotometer to calculate the cumulative amount of released canthaxanthin.
Each experiment was carried out in triplicate and shown as mean ± standard deviation (SD).

Fish and Diet.
e effects of canthaxanthin-and α-tocopherol-loaded liposomes as diet supplemented for feeding rainbow trout were investigated at Sapa Research Center for Coldwater Aquaculture, Research Institute for Aquaculture (Lao Cai Province, Vietnam). A total of 900 fish were individually weighed and randomly distributed into 30 tanks (280 L each). Each tank contained 30 fish, and each replication comprised 10 tanks. One month prior to dietary supplementation, fish were raised on noncolored feeds to have their muscle color returned to original color. Afterwards, each tank in each replication group was fed with one of the ten experimental diets for another 3 months (Table 1). e dietary feeds used in the experiment were produced by Kinh Bac Feed Mill, Bac Ninh province, Vietnam, with the basic nutrient content as follows: total protein 40%, total fat 18%, crude fiber 5%, and energy of 3800 kcal/kg. Fish were fed at 6 am, 10 am, 2 pm, and 6 pm every day until full. Tank water was continuously replaced with fresh water at the rate of 4 L/min. Water temperature and dissolved oxygen were maintained in the range from 6 to 17.5°C and from 10 to 11 ppm, respectively. Air pumps were used for water aeration.

Weight Measurement.
During the trial, fish were periodically weighed every month by taking the mean weight of three randomly caught fish from each treatment group. Growth indicators of fish were determined as follows: Here, W f is the final body weight (g), W i is the initial body weight (g), and t is the experimental duration (days).

Muscle Canthaxanthin Extraction and Determination.
After three months of feeding trial, fish fillets were collected and measured for color using SalmoFan Lineal colorimeter (DSM Nutritional Products, Kaiseraugst, Switzerland) on a scale of 20 to 34, which is the accepted international benchmark for salmon color (Figure 2). Scores recorded for each animal were made in agreement with three researchers.
Canthaxanthin disposition in fish muscle tissue was determined as follows. First, muscle samples were collected on the fish dorsal and stored at −20°C. en, canthaxanthin was extracted following a previously described procedure [16]. Briefly, after crushing with an IKA basic ULTRA-TURRAX ® homogenizer, the samples were extracted with organic solvents, then cleaned, and analyzed by a HPLC chromatograph with the standard canthaxanthin of Sigma Aldrich. e liquid gas chromatography system used the C- 18 Hypersil Gold analysis column (5 µm, 150 mm × 4.6 mm), with flow rate set to 1 mL/min. e retention time was 6 minutes. e process was carried out at the analysis room of the Institute of Chemistry and Natural Compounds.

Statistical Analysis.
Experimental data was calculated with mean and standard error. Average data of the treatments was processed by ANOVA on Minitab 16. Duncan comparison was used to differentiate each treatment. e difference is considered significant when p < 0.05. e canthaxanthin spectrum data are given in Table 2. Total amount of canthaxanthin from 1 g dry biomass was 1.5 mg. Yield � 1.5/1000 × 100 � 0.15%.

Synthesis and Characterization of Canthaxanthin-and a-Tocopherol-Loaded
Liposomes. Liposomes were made by the established thin-film hydration method followed by extrusion. e physicochemical properties of the optimal blank lecithin liposome, canthaxanthin-loaded liposomes, α-tocopherol-loaded liposomes, and canthaxanthin-and α-tocopherol-loaded liposomes formulation were studied. As shown in Figure 3, the mean size of blank lecithin liposome was 105.53 ± 9.02 (mean ± SD; n � 3), while the mean sizes of canthaxanthin-loaded liposomes and α-tocopherol-loaded liposomes were 97.33 ± 4.64 and    103.80 ± 6.95 (mean ± SD; n � 3), respectively. In case of canthaxanthin-and α-tocopherol-loaded liposomes, the mean sizes of liposomes with canthaxanthin IC � 0.1%, IC � 0.5%, and IC � 1% were 109.70 ± 6.36, 105.10 ± 8.41, and 109.20 ± 5.66 (mean ± SD; n � 3). Polydispersity index (PDI), an indicator of dispersion homogeneity, of all formulated liposomes ranged from 0 and 0.2, suggesting that active ingredients were homogenously dispersed on the synthesized liposomes. PDI value of α-tocopherolloaded liposomes was lowest at 0.150 ± 0.044 (mean ± SD; n � 3). All other liposomes have higher PDI values but remain lower than 0.2. ere were no significant differences of the mean size and PDI between blank lecithin liposome and canthaxanthin-and α-tocopherol-loaded liposomes. e encapsulation efficiency (EE%) and drug loading content (DL %) of canthaxanthin-loaded liposomes were 59.6 ± 2.3% (mean ± SD; n � 3) and 2.39 ± 0.23% (mean ± SD; n � 3), respectively (Table 3). e encapsulation efficiency of canthaxanthin-and α-tocopherol-loaded liposomes decreased when increasing the concentration of canthaxanthin in lecithin liposome. EE% values of liposomes loaded with canthaxanthin at IC � 0.1%, IC � 0.5%, and IC � 1% were 85.3 ± 2.1, 72.9 ± 1.8, and 55.3 ± 2.6, respectively. e role of lipid layer in protecting core materials of the liposome is accentuated by the fact that active ingredients such as carotenoids easily degrade under acidic microenvironments and by the influence of enzymes. In vitro release profiles of canthaxanthin from canthaxanthin-loaded liposomes, canthaxanthin-and α-tocopherol-loaded liposomes with IC � 0.1%, canthaxanthin-and α-tocopherol-loaded liposomes with IC � 0.5%, and canthaxanthin-and α-tocopherol-loaded liposomes with IC � 1% in PBS at pH 7.4 are shown in Figure 4. Within the first 4 hours, release of canthaxanthin from liposomes formed without α-tocopherol seemed to be more rapid than other liposomes. By contrast, liposomes loaded with canthaxanthin at IC � 0.1% showed slower, sustained release rate, which reached 0.46 after 4 h. In addition, increasing the loaded canthaxanthin concentration apparently induced faster canthaxanthin release within the first 1 h. After 4 h of incubation, approximately 68% of the encapsulated canthaxanthin was released from the liposome containing 1% canthaxanthin. ese data suggest that canthaxanthin can be well encapsulated in liposomes and released in an extended period.

Fish Growth.
e culture of rainbow trout was carried out in the winter to ensure suitable growth conditions and water temperature [8]. Dissolved oxygen in experiment was controlled and ranged from 7.1 to 10.9 mg/l, with an average of 8.2 mg/L [28,29].
Initially, weight of the fish was 309.43 ± 12.98 g before being raised with feeds, and there was no difference between groups. Mean wet weights of fish were recorded monthly during feeding trial and once before filleting. During the experiment, the fish remained in good health, and the mortality rate was 0%. At the end of the feeding trial, fish fed with diet supplemented with 1 g/kg canthaxanthin-and α-tocopherol-loaded liposomes (IC � 0.5%) showed significantly higher average weight than those of other experimental control groups (p < 0.05) ( Figure 5).
Rainbow trout fed with diet supplemented with canthaxanthin at 20 mg/kg (diet IV) showed significantly lower wet weights (p < 0.05) after two, three, and four months compared to other groups. is finding is in line with that reported by Torrissen et al. [17] where increased dose of fed canthaxanthin resulted in the lower weight gain in cultured rainbow trout. Final weights of fish in groups fed with diets supplemented with 1 g/kg lecithin or 100 mg/kg α-tocopherol were higher, albeit not statistically significant (p > 0.05), than those fed with unsupplemented commercial diet. In case of groups fed with diets supplemented with canthaxanthin-and α-tocopherol-loaded liposomes (diets VII, IX, and X), their weight gain was higher when increasing the concentration of canthaxanthin from 0.1% to 0.5% but was stagnant when further increasing concentration of canthaxanthin from 0.5 to 1% (not significant difference).
Current results are similar to those of previous reports pointing out that inclusion of carotenoids such as canthaxanthin or astaxanthin in fish feed affected the growth rate and efficiency of food usage of freshwater salmon or parrot fish [30,31]. In addition, several studies also reported that carotenoids could enhance growth by acting as fertilization hormones, but, up to date, biological functions of the carotenoids in fish have not been unconfirmed [32,33]. Table 4 shows the results of the muscle proximate composition of rainbow trout after 3 months of feeding with various experimental diets. No significant differences were recognized among 10 treatment groups in all indicators (crude protein, crude lipid, ash, and moisture). ese findings are not consistent with previous reports demonstrating that dietary canthaxanthin deficiency can affect digestion and absorption of lipid. To be specific, it was found that dietary supplementation of carotenoids did affect digestion and absorption of lipid in rats [34]. Similarly, for aquatic animals, Brizio et al. noted that canthaxanthin affected the digestion of lipid in fish [35]. Nevertheless, the effects of canthaxanthin and α-tocopherol on fish muscle composition are still controversial and complex, and further studies are needed to reveal their mechanisms.

Muscle Canthaxanthin.
At the beginning of the experiment, the color of fish muscle was uniform among experimental groups with the color score ranging from 20.1 to 20.9 (Table 5). After 1 month of feeding, the color began to show differences between treatments (p < 0.05). e highest score (27.0) was recorded in fillets of fish fed with commercial diet supplemented with 100 mg/kg α-tocopherol and 20 mg/kg canthaxanthin (diet V), closely followed by that of diet IV and diet IX (26.9). e lowest score of 23.1 was recorded in diet I, which is in absence of canthaxanthin. After 2 months, the difference in color of the muscle became more apparent ( Figure 6). However, towards the end of the experiment, improvements in color score were less noticeable, which is consistent with previous studies.
is result is similar to previous research with the astaxanthin and canthaxanthin supplementation at the ratio of 40 : 40 mg/kg of food [36]. In addition, Choubert et al. and Torrissen et al. also reported that the color of the fish muscle would gradually not increase and reach saturation, even with increased concentration or longer feeding [10,17].
e concentration of canthaxanthin accumulated in fish muscle in different treatments after 90 days of culturing is shown in Figure 7. After 90 days of feeding, fish fed with different experimental diets showed varying concentrations of canthaxanthin accumulated in the fish muscle (p < 0.05). Canthaxanthin content in muscle was highest (2.91 mg/kg)  in the group fed with diet supplemented with 1 g/kg canthaxanthin-and α-tocopherol-loaded liposomes (IC � 0.5%), followed by diet supplemented with 1 g/kg canthaxanthin-and α-tocopherol-loaded liposomes (IC � 1%) (2.90 mg/kg) and commercial diet supplemented with 100 mg/kg α-tocopherol and 20 mg/kg canthaxanthin (2.75 mg/kg). e lowest canthaxanthin content in muscle was found in the group fed with commercial diet supplemented with 1 g/kg lecithin (0.99 mg/kg). However, between diet IX and diet V, there is no statistically significant difference in canthaxanthin disposition.
Economic viability of diet IX (1 g/kg canthaxanthin-and α-tocopherol-loaded liposomes, IC � 0.5%) is also justified by the prices of canthaxanthin and α-tocopherol. To be    [37]. is suggests that the cost of canthaxanthin in the composition of diet IX is approximately four times lower than those of commercial fish feed supplemented with 20 mg/kg of unencapsulated canthaxanthin and that formulation of liposomal fish diet carries important implications for reducing carotenoid feeding while still meeting international standards. According to Torrissen et al., canthaxanthin disposition in rainbow trout muscle was proportional to the ratio of additional canthaxanthin in food [17]. However, with the inclusion of canthaxanthin at a concentrations higher than 50 mg/kg, the accumulation of canthaxanthin in the rainbow trout muscle decreased gradually and stopped when reaching saturation.
Interestingly, canthaxanthin deposition in fish flesh increased in the presence of α-tocopherol in the feed (diet V), which is in line with findings reported by Choubert et al. [10] proposing that, compared to feeding using bare canthaxanthin, the use of liposomes formed from lecithin and α-tocopherol could reduce the required amount of canthaxanthin needed to achieve the same color score. Further experiments are needed to clarify this assertion.

Conclusions
In the present study, liposomes containing α-tocopherol and canthaxanthin as active ingredients were fabricated and included in fish feeds of cultured rainbow trout. ere was no difference in growth rate, survival rate, or flesh nutritional composition between cultured rainbow trout fed with diets supplemented with bare lecithin, α-tocopherol, and canthaxanthin and those fed with liposomes formed from lecithin, α-tocopherol, and canthaxanthin. However, the color and canthaxanthin disposition in fish muscle of fish fed with diets containing 1 g/kg canthaxanthin-and α-tocopherol-loaded liposomes (IC � 0.5%) were similar to those of fish fed with commercial diet supplemented with 20 mg/kg of unencapsulated canthaxanthin. However, the experimental diet containing liposomal canthaxanthin and α-tocopherol only required half of the amount of canthaxanthin compared to the diet supplemented with bare canthaxanthin. ese results suggest the feasibility of using liposome formation techniques in manufacture of fish diet supplements. Further studies should focus on mechanisms of canthaxanthin pigmentation on fish flesh and evaluate industrial-scale utilization of supplemented fish diets.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare no conflicts of interest.