Effect of an Emulsified Formulation on Vegetable Carotenoid Bioaccessibility

vitro digestion assays ％（ w/v ） used of 0.3 ％ – concentrations in previous were more than twice as high as those reported 4 − 11 ） . This high concentration may have been the main reason why the effects of fats and oils on lutein bioaccessibility were not reported. In this study, bile extract and pancreatin concentrations in an in vitro digestion assay were adjusted to 2.5 ％（ w/v ） and 0.4 ％（ w/v ） , respectively, and the effects of various lipids, including polar lipids, on vegetable carotenoid bioaccessi-NOTE Abstract: We investigated the effect of neutral lipids, polar lipids, and an emulsified formulation (EMF) on carotenoid bioaccessibility in an in vitro digestion assay of vegetables. These reagents enhanced carotenoid bioaccessibility. Contrary to our previous report, they also exhibited effects on lutein. Bile extracts/ pancreatin concentrations also participated in the bioaccessibility. The EMF, which consisted of lower amounts of oil, had the same effect on lutein as rapeseed oil. These reagents also showed effects in the aging model, with more reduced bile extract/pancreatin concentrations, suggesting that lipids and EMF contributed to carotenoid bioaccessibility in bile/pancreatic juice secretions due to aging and disease.

words, concentrations in our previous study were more than twice as high as those reported 4 11 . This high concentration may have been the main reason why the effects of fats and oils on lutein bioaccessibility were not reported. In this study, bile extract and pancreatin concentrations in an in vitro digestion assay were adjusted to 2. 5 w/v and 0.4 w/v , respectively, and the effects of various lipids, including polar lipids, on vegetable carotenoid bioaccessi-bility were examined.
Although it was recommended that fats and oils are consumed together to increase carotenoid absorption, fat and oil consumption is also of concern in terms of increased calorie intake. If some neutral lipids could be replaced with polar lipids or sugar alcohols, the same weight with lower calories may produce the same or better effect as neutral lipids. We also examined the effect of an emulsified formulation EMF , which had the same weight and lower calorie content when compared with neutral lipids Table 1 . If bile and pancreatic juice secretion is reduced because of aging or disease, the bioavailability of fat-soluble components will be reduced 12 14 . We simulated such a model with a much lower concentration of bile and pancreatic juice and investigated the effect of EMF on hyposecretion.
The main objective of this study was to determine the effects of EMF in comparison to the effects of various lipids on the carotenoid bioaccessibility.

Materials
Soybean phosphatidylcholine PC , soybean lysophosphatidylcholine lysoPC , pepsin from porcine gastric mucosa, pancreatin from porcine pancreas, and porcine bile extract were purchased from Sigma-Aldrich St. Louis, MO, USA . Pyrogallol was purchased from FUJIFILM Wako Pure Chemical Corporation Osaka, Japan . Egg PC PC-98N and egg lysoPC LPC-1 were obtained from Kewpie Co. Tokyo, Japan . Soybean oil was purchased from Clea Japan Inc. Tokyo, Japan . Rapeseed oil was purchased from MP Biomedicals Santa Ana, CA, USA . These oils were purified as outlined in our previous report 15 . Medium-chain fatty acid triglyceride MCT oil was purchased from the Nisshin OilliO Group Tokyo, Japan . β-Carotene and lutein standards were prepared as previously described 16 . All other chemicals and solvents were of reagent grade.

Preparation of vegetable homogenates
Spinach Spinicia oleracea and carrot Daucus carota were purchased at a supermarket in Tsukuba City, Japan. Homogenates were prepared using a modification of our previous method 3 . Briefly, spinach roots and carrot peel were discarded and the edible components were boiled and cooled at room temperature. Water was then added at a ratio of 1:2 by weight. This water contained pyrogallol at 10 of the vegetable weight. The mixture was homogenized on ice using a Polytron blender PT2500E, Kinematica AG, Malters, Switzerland .

Carotenoid content in vegetable homogenates
Carotenoids were first extracted from 1.5 g homogenate using 1.5 mL ethanol, 1.5 mL ethyl acetate, and 1.5 mL hexane. For second and third extractions, 1.5 mL ethyl acetate and 1.5 mL hexane were used. The organic solvent fractions were pooled, dried in a centrifugal evaporator, and subjected to high-performance liquid chromatography HPLC analysis see below .

In vitro digestion assay
Gastrointestinal digestion was simulated on the basis of our previous report, with modifications 3 . To investigate the effect of neutral lipids, polar lipids, and EMF on vegetable carotenoid solubilization during digestion, these reagents were added to homogenates and mixed well. The mixture was mixed with 3 mL 0.5 pepsin in a buffer 3 . The pH was adjusted to 2.0 with HCl, and then, the mixture was shaken at 120 rpm in the dark at 37 for 1 h under nitrogen. After incubation, the pH was increased to 5.0 with NaHCO 3 and 3 mL of pancreatin and bile extract in 0.1 mol/ L NaHCO 3 was added. After the pH was further increased to 7.5 with NaOH, the mixture was shaken at 120 rpm in the dark at 37 for 2 h under nitrogen. Afterward, the digest was centrifuged at 2900 g for 20 min at 4 , and part of the supernatant was passed through a 0.2 μm filter to generate a solubilized carotenoid fraction. We regarded this filtrate as a solubilized fraction. Bioaccessibility was defined as the ratio of carotenoid in the filtrate solubilized fraction to carotenoid in the vegetable homogenate. For lutein analysis in spinach, the filtrate was diluted in nine volumes of dichloromethane: methanol 1:4, v/v . An aliquot was then subjected to HPLC. Since the amount of β-carotene solubilized from carrots was low, carrot β-carotene was extracted from the filtrate as previously described 16 and subjected to HPLC.

HPLC analysis
Carotenoids were analyzed using HPLC, as previously described 16 . However, LC-10AD and SPD-10A UV-VIS were used for the pump and detector, respectively.

Statistical analysis
Data were processed using a one-way analysis of variance, followed by the Tukey-Kramer method. A p-value of 0.05 was considered statistically significant.

Results
3.1 Effect of neutral lipids, polar lipids, and EMF on lutein bioaccessibility from spinaches The effect of oils and fats on lutein bioaccessibility from spinaches were investigated using the neutral lipids, namely, soybean oil, rapeseed oil, and MCT oil Fig. 2A , and the polar lipids, namely, soybean PC, soybean lysoPC, egg PC, and egg lysoPC Fig. 2B , and were compared with EMF effects.
In these lipids or EMF, the effect on bioaccessibility tended to increase in a concentration-dependent manner. We observed no statistical differences in effects among oils or phospholipids at 15 mg. Similarly, no differences were observed between those lipids and EMF. The addition of 15 mg rapeseed oil or EMF increased lutein bioaccessibility approximately 1.5 times when compared with no addition 0 mg .
3.2 Effect of rapeseed oil and EMF on β-carotene bioaccessibility from carrots Of the lipids tested in Fig. 2A, rapeseed oil tended to exhibit the highest effect on the bioaccessibility, although this was not statistically different. Thus, the effects of rapeseed oil and EMF on β-carotene bioaccessibility from carrots were examined and compared. As shown Fig. 3 , the addition of 15 mg rapeseed oil or EMF increased bioaccessibility by approximately four or two times, respectively.

Effect of bile extract and pancreatin concentrations
on carotenoid bioaccessibility from vegetables The effect of bile extract and pancreatin concentrations on the in vitro digestion of vegetable carotenoids was studied. Neither lipids nor EMF was added to the assay. Both spinach lutein Fig. 4A and carrot β-carotene Fig.  4B bioaccessibility decreased significantly with decreasing bile extract and pancreatin concentrations.

Effect of rapeseed oil and EMF on vegetable carotenoid bioaccessibility in a decreased bile and pancreatin secretion model
The effects of rapeseed oil and EMF when bile extract and pancreatin concentrations in the in vitro digestion assay were reduced to one-tenth of the concentration Figs. 2 and 3 , i.e., from 2.5 to 0.25 and 0.4 to 0.04 , respectively, are shown Fig. 5 . Both showed a concentration-dependent increase in spinach lutein Fig.  5A and carrot β-carotene bioaccessibility Fig. 5B . We observed no differences in effects between rapeseed oil and EMF at all concentrations.

Fig. 2 Effect of various lipids and emulsified formulation
EMF on lutein bioaccessibility in spinach. Neutral lipids A and phospholipids B at 0-15 mg were added to spinach homogenates. EMF data were the same for A and B . Bile extract and pancreatin concentrations for in vitro digestion assays were 2.5 and 0.4 , respectively. Bars represent the mean standard deviation SD n 4 . Values not sharing a common letter are significantly different by the Tukey-Kramer test p 0.05 .

Discussion
Previously, neutral lipids exerted no effects on increasing lutein bioaccessibility in vegetables 3 . However, these lipids showed effects when bile extract and pancreatin concentrations were reduced in the in vitro digestion assay Fig. 2A , suggesting the effects of neutral lipids on lutein solubilization during digestion after vegetable consumption are influenced by bile and pancreatic juice secretion. Since the concentration of bile in this study was half that in the previous report, the ability to solubilize lutein was reduced, but the reduction was reinforced by free fatty acids, which are hydrolyzed products of lipids. Therefore, the effect of lipids would have been apparent.
The fatty acid composition of the rapeseed oil used in this study was C16:0 4 , C18:0 1 , C18:1 13 , C18:2 15 , C18:3 16 , and C22:1 51 according to the product catalog. We previously showed the effect of various free fatty acids on vegetable carotenoid bioaccessibility 3 , indicating that the fatty acid type produced by lipid hydrolysis affected carotenoid bioaccessibility. C18:1 was the most effective but did not differ from C22:1 3 . The elevated effect of rapeseed oil may have been attributed to the fact that C18:1 and C22:1 were the major constituent fatty acids in rapeseed oil. Since C18:1 is the major fatty acid in EMF, this may have been a factor contributing to its elevated effects. However, the total composition 64 of C18:1 and C22:1 in rapeseed oil was almost the same as that of C18:1 62.4 in EMF. Hence, the fatty acid composition was different between the two; however, this difference in chain length of fatty acids would not affect their differential effects on bioaccessibility. The lysophospholipid that is hydrolyzed products of phospholipid may also affect the effect of phospholipids on bioaccessibility Fig. 2B . The effect of EMF on β-carotene bioaccessibility was lower than rapeseed oil Fig. 3 . However, EMF has approximately 30 fewer calories when compared with the same number of lipids Table 1 . If a comparison was made at the same calorie level 10 mg rapeseed oil, which is 0.90 kcal, is about equivalent to 15 mg EMF, which is 0.93 kcal , then EMF would have had a same effect Fig. 3 .
In the elderly, bile and pancreatic juice secretion is reportedly reduced 12 . Vegetable carotenoid bioaccessibility was decreased when bile extract and pancreatin concentrations in the in vitro digestion assay were lowered Fig.  4 . We hypothesized this decrease may have been compensated by lipids and EMF Fig. 5 . However, the low pancreatin concentration may decrease the lipid hydrolysis rate, which reduces the effects of promoting carotenoid solubilization. Thus, rapeseed oil may not be sufficiently effective on the bioaccessibility. Alternatively, sugar alcohols/dextrin and phospholipids in egg yolk oil 17 , which are EMF components Table 1 , may have worked to promote solubilization even under such conditions. This mechanism requires elucidation in the future.  The intake of oils increases concerns regarding increased calorie intake; however, EMF generated the same level of carotenoid bioaccessibility at a lower calorie level when compared with vegetable oils. Alternatively, issues for the elderly include low nutrition and calorie intake rather than excessive calorie intake 18 . In other words, a high intake of oil may be preferable, but a decrease in digestive enzyme secretion due to aging reduces carotenoid bioaccessibility. Thus, EMF consumption would be more favorable; since EMF has 30 fewer calories than vegetable oils of the same weight, there is room to add another calorie as a constituent oil of EMF in a calorie-based comparison. If this could be done, it would increase the potential of EMF in the bioaccessibility.
MCT oil reportedly increases body weight by increasing muscle mass when compared with salad oil containing the same calories 19 . Replacing the oil component of EMF with MCT oil may help EMF prevent frailty in the elderly. These hypotheses must be confirmed in future human subject trials.

Conclusions
Bile extract and pancreatin concentrations in our in vitro digestion assay significantly affected vegetable carotenoid bioaccessibility. In our previous report, neutral lipids did not promote lutein bioaccessibility. This was because the bile concentration was too high at 5 . Here various lipids showed effects on the bioaccessibility. EMF, comprising smaller amounts of oil also had similar effects on lutein as rapeseed oil, whereas rapeseed oil had a higher effect on β-carotene. Rapeseed oil and EMF also showed the same effects in the aging model, with reduced bile extract and pancreatin concentrations, suggesting that these molecules exert compensatory effects on the decrease in bile/pancreatic juice secretions due to aging and disease. EMF that can serve calories is may be applied to nursing care and post-sickness meals.