Absorption Kinetics of Ethanolamine Plasmalogen and Its Hydrolysate in Mice.

Ethanolamine plasmalogen (PlsEtn), a subclass of ethanolamine glycerophospholipid (EtnGpl), has been reported to have many biological and dietary functions. In terms of PlsEtn absorption, some studies have reported that PlsEtn is re-esterized at the sn-2 position using lymph cannulation and the everted jejunal sac model. In this study, we aimed to better understand the uptake kinetics of PlsEtn and increase its absorption. We thus compared the uptake kinetics of PlsEtn with that of the lyso-form, in which the fatty acid at the sn-2 position was hydrolyzed enzymatically. Upon administration of EtnGpl (extracted from oysters or ascidians, 75.4 mol% and 88.4 mol% of PlsEtn ratio, respectively), the plasma PlsEtn species in mice showed the highest levels at 4 or 8 hours after administration. In the contrast, administration of the EtnGpl hydrolysate, which contained lysoEtnGpl and free fatty acids, markedly increased the plasma levels of PlsEtn species at 2 h after administration. The area under the plasma concentration-time curve (AUC), especially the AUC0-4 h of PlsEtn species, was higher with hydrolysate administration than that with EtnGpl administration. These results indicate that EtnGpl hydrolysis accelerated the absorption and metabolism of PlsEtn. Consequently, using a different experimental approach from that used in previous studies, we reconfirmed that PlsEtn species were absorbed via hydrolysis at the sn-2 position, suggesting that hydrolysis in advance could increase PlsEtn uptake.

specific diseases. PlsEtn levels in the brain and blood are lower in patients with Alzheimer s disease and in those with other neurological diseases compared to the levels in healthy subjects 1,6 . PlsEtn metabolism is also profoundly dysregulated in dedifferentiated colon mucosa 7 .
In contrast, some marine invertebrates e.g., ascidians, oysters, and mussels are known to be rich in PlsEtn along with DHA and EPA 8,9 , and their dietary functions are rather interesting. Dietary PlsEtn from ascidians improves impaired memory in model rats of Alzheimer s disease 10 . PlsEtn, especially species bearing DHA, suppress neuronal apoptosis in vitro 11 . Dietary PlsEtn from ascidians suppresses the formation of aberrant crypt foci and colon inflammation in colon cancer 12 .
To effectively utilize dietary functions of PlsEtn, it is necessary to understand its digestion, absorption, and metabolism. In terms of PlsEtn absorption, Hara et al. reported the presence of a small but significant amount of PlsEtn in the lymph but not in the portal vein of rats after duodenal infusion of a phospholipid fraction containing PlsEtn, thereby indicating that PlsEtn was absorbed in vivo 13,14 . Interestingly, Nishimukai et al. have reported that the composition of PlsEtn differs between the administered lipid emulsion and the collected lymph, thus implying structural changes in PlsEtn i.e., preferential AA re-esterification at the sn-2 position and base conversion of PlsEtn to choline plasmalogen PlsCho during absorption 15 . Previously, we also used lymph cannulation and the everted jejunal sac model to confirm that upon administration, PlsEtn was preferentially re-esterized to AA and partly converted to choline base 16 .
In light of the aforementioned studies, we compared the uptake kinetics of PlsEtn with that of the lyso-form in the plasma of mice to better understand the uptake kinetics of PlsEtn and to increase the absorption of PlsEtn. EtnGpl containing high levels of PlsEtn was prepared from oyster or ascidian muscle, and lysoEtnGpl with a high level of ly-soPlsEtn was obtained from EtnGpl by enzymatic hydrolysis at the sn-2 position.

Materials and reagents
Japanese oysters Crassostrea gigas were purchased from a local fish market in Hokkaido and were lyophilized within one day of purchasing and crushed. Freeze-dried ascidian Halocynthia roretzi muscle was provided by Yaizu Suisankagaku Industry Co., Ltd. Shizuoka, Japan . Standard phospholipid species were purchased from Avanti Polar Lipids, Inc. Alabaster, AL, USA , and 18:0ol/20:5-PlsEtn was purified according to previously reported methods 17 . All other reagents used were purchased from FUJIFILM Wako Pure Chemical Co. Osaka, Japan unless stated otherwise.

Preparation of EtnGpl and the lyso-form from oysters
and ascidians EtnGpl was prepared from oysters or ascidian muscle as per modification of our previously reported method 17 . Ly-soEtnGpl was hydrolyzed from EtnGpl using Brockehoff s method 18 . EtnGpl was purified using silica column chromatography, and the purified EtnGpl was hydrolyzed with phospholipase A 2 PLA 2 from Crotalus atrox venom Sigma-Aldrich, St. Louis, USA in 0.22 M NaCl, 20 mM CaCl 2 , 1 mM EDTA-2Na, and 0.05 M MOPS pH 7.2 at 30 . After the reaction, the lipophilic fraction containing lysoEtnGpl and free fatty acids FFA was obtained by solvent fractionation.

Animals and plasma preparation
Male ICR mice were obtained from CLEA Japan, Inc. Tokyo, Japan at 7 weeks of age and were housed under pathogen-free conditions in micro-isolator cages at 22 1 under a 12-h light/12-h dark cycle. The mice were acclimated for 1 week with CE-2 diets CLEA Japan, Inc. and then used for the absorption experiments. After a 12-h starvation period, the following samples were administered into the stomach by oral gavage: for Experiment 1, the control group was administered with 0.5 mL of 5 NaHCO 3 and 1 sodium taurocholate as the vehicle, the EtnGpl group was administered with 0.5 mL of oyster EtnGpl 390 μmol/kg body weight sonication-emulsified in the vehicle, and the LysoEtnGpl group was administered oyster Ly-soEtnGpl 390 μmol/kg body weight plus FFA emulsified in vehicle; for Experiment 2, the control group was administered with 0.5 mL of 5 NaHCO 3 , 6.7 sodium taurocholate, and 1.67 fatty acid-free bovine serum albumin as the vehicle, the EtnGpl group was administered with 0.5 mL of ascidian EtnGpl 390 μmol/kg body weight sonication-emulsified in the vehicle, the LysoEtnGpl group was administered with ascidian LysoEtnGpl 390 μmol/kg body weight emulsified in the vehicle. Meanwhile, the Mix group was administered with a mixture of ascidian EtnGpl and LysoEtnGpl 195 μmol/kg body weight emulsified in the vehicle. The mice were then anesthetized with sodium pentobarbital 65 mg/kg BW, i.p. at the scheduled time. Fresh blood samples were freshly collected from the right ventricle in tubes with EDTA-2Na, and subjected to lowspeed centrifugation 15 min, 1,000 g, 4 for plasma separation. The plasma was stored at 80 until use. All animal experiments were performed with protocols approved by the Animal Care and Use Committee and according to the Obihiro University Guidelines Permit Number: 28-191 and 18-180 .

Lipid analysis
Fatty acid methylester and aldehyde composition were determined by gas chromatography 19

Statistical analysis
Differences among all data groups were assessed using one-way ANOVA with the Fisher s least significant difference test; P values 0.05 were considered as statistically significant. All data were analyzed using BellCurve for Excel Social Survey Research Information Co., Ltd., Tokyo, Japan and were expressed as mean standard error of the mean SEM .

Lipid composition of EtnGpl and the hydrolysate from
oysters and ascidians TLC analysis confirmed that the enzymatic reaction generated lysoEtnGpl and FFA from EtnGpl Fig. 1 . The alkenyl moiety in oyster EtnGpl mostly comprised 18:0ol, and the prominent acyl moieties were EPA and DHA, while the ratio of AA was much lower Table 1 . The composition of the alkenyl and acyl moieties in the hydrolysate was nearly the same. In oyster EtnGpl, PlsEtn-18:0ol/22:6 was the predominant PL species, PlsEtn-18:0ol/20:4 was a minor species, and ChoGpl species were not observed. In the hydrolysate, only lysoEtnGpl species were detected, and the most prominent PL species was lysoPlsEtn-18:0ol.
The alkenyl moiety in ascidian EtnGpl mostly comprised 18:0ol, and the prominent acyl moiety was EPA, whereas AA was a minor moiety Table 2 . The composition of the alkenyl and acyl moieties in the hydrolysate was almost the same. PlsEtn-18:0ol/20:5 was also analyzed in ascidian EtnGpl because 18:0ol and EPA were abundant. In ascidian EtnGpl, PlsEtn-18:0ol/20:5 was the prominent PL species and PlsEtn-18:0ol/20:4 was a minor species. In the hydrolysate, lysoEtnGpl species, especially lysoPlsEtn-18:0ol, was the most abundant and PlsEtn-18:0ol/20:5 was observed at a lower ratio. The PlsEtn ratio of oyster and ascidian EtnGpl was found to be as 75.4 mol and 88.4 mol as per each carbon chain composition, respectively.

Effects of administration of oyster EtnGpl and its hy-
drolysate on the carbon chain composition and PL species levels in plasma Experiment 1: Chronological changes in the carbon chain composition and PL species levels in plasma were investigated after administration of oyster EtnGpl and its hydrolysate. Administration of oyster EtnGpl increased the 18:0ol ratio in the plasma, which was composed of plasmalogens, at 4 h and DHA ratio in the plasma at 1 and 4 h after administration compared to the control group Fig. S1 . Administration of the hydrolysate also increased the EPA ratio in the plasma at 1 and 2 h and decreased the AA ratio at 2 h. Furthermore, administration of the hydrolysate increased the 18:0ol ratio in the plasma at 2 h compared to that in the control and EtnGpl groups.
When compared to the control group, administration of oyster EtnGpl significantly increased the plasma levels of PlsEtn-18:0ol/18:1 and 18:0ol/20:4 at 2 and 4 h after administration, and PlsEtn-18:0ol/22:6 at 4 h Fig. 2 . Administration of the hydrolysate also markedly increased the plasma levels of PlsEtn-18:0ol/20:4 at 1 h and PlsEtn-18:0ol/22:6 at 2 h. Notably, the LysoEtnGpl group showed importantly higher levels of PlsEtn-18:0ol/20:4 and 18:0ol/22:6 at 2 h when compared to the EtnGpl group. The LysoEtnGpl group showed the highest levels of lysoPlsEtn-18:0ol at 1 and 2 h among the groups, and a significantly higher level at 4 h compared to the control group, but a meaningfully lower level than that of the EtnGpl group. The EtnGpl   In terms of ChoGpl species that were not detected in the administered samples, when compared to the control group, the LysoEtnGpl group showed importantly higher levels of PlsCho-18:0ol/18:1 at 1, 2, and 4 h, PlsCho-18:0ol/20:4 at 2 and 4 h, and PlsCho-18:0ol/22:6 at 2 h, whereas the EtnGpl group showed higher levels of PlsCho-18:0ol/18:1 and 18:0ol/20:4, but these were not significant. The plasma levels of other ChoGpl species were not significantly different among the groups.
3.3 Effects of administration of ascidian EtnGpl, its hydrolysate, and their mixture on carbon chain composition and PL species levels in the plasma Experiment 2: To further confirm and clarify the PlsEtn kinetics, we analyzed the plasma data until 8 h after administering ascidian EtnGpl, its hydrolysate, and the mixture with an equivalent amount of total PL. The EtnGpl source and auxiliary composition of the emulsion for this experiment was different from those in Experiment 1. Albumin was added and the concentration of sodium taurocholate increased. This was done to reduce the possibility of emulsion formation by lysoPL, which forms an emulsion more easily than PL.
In terms of carbon chain composition in the plasma, when compared with the control group, administration of ascidian EtnGpl significantly increased the plasma EPA ratio at 1, 2, and 4 h after administration and did not affect the other ratios Fig. S2 . Administration of the hydrolysate also markedly increased the plasma ratio of 18:0ol at 4 h. Additionally, administration of the hydrolysate significantly increased the plasma EPA ratio at 1 h compared to that in the EtnGpl group.
When compared to the control group, administration of ascidian EtnGpl significantly increased the plasma levels of In terms of ChoGpl species, the LysoEtnGpl group showed significantly higher levels of PlsCho-18:0ol/20:4 at 1, 2, and 4 h, of PlsCho-18:0ol/22:6 at 4 h and higher levels of lysoPlsCho-18:0ol at 2 and 4 h. At 8 h after administration, the plasma levels of PlsCho-18:0ol/20:4 and 18:0ol/22:6 and lysoPlsCho-18:0ol in the LysoEtnGpl group were meaningfully decreased and were almost the same as those in the control group, whereas these levels in the EtnGpl group increased and were the highest among all groups. Plasma kinetics of PlsCho species in the Mix group tended to exhibit intermediate kinetics between those of EtnGpl and LysoEtnGpl groups. Plasma levels of other ChoGpl species were not significantly different among the groups.

Effects of administration of ascidian EtnGpl, its hydrolysate, and their mixture on the AUC of PL species
The area under the plasma concentration-time curve AUC of each PL species was calculated by subtracting the values in the control group based on Fig. 3 Fig. 4 . As plasma was obtained from different mice at different times, average values were used and could not be tested for significant differences.
Clarified increases in AUC were observed in PL species related to the alkenyl chain 18:0ol, which was predominant in the administered samples from ascidians. Hydrolysis of EtnGpl raised the AUC of these species, especially in the early stage compared to that of untreated EtnGpl. The mixture of untreated and treated EtnGpl showed intermediate AUC levels.

Discussion
PlsEtn has been reported to exert many biological functions on various parts of the body. These functionalities include improvement of cognitive impairment, suppression of aberrant crypt foci formation and inflammation in the colon, suppression of cholesterol synthesis in in vitro cells, suppression of atherosclerosis formation, and alleviation of lead-cytotoxicity in in vitro liver cells 9 12, 21 23 . As it is important to understand PlsEtn absorption and metabolism to clarify its function, some studies on PlsEtn absorption have been performed using lymph cannulation and everted jejunal sac model. These studies indicate that the administered PlsEtn is preferentially re-esterized to AA and partly converted to PlsCho during absorption 15,16 . In the present study, we analyzed the kinetics from a different perspective and compared the uptake kinetics of PlsEtn with the lyso-form in the plasma of mice. EtnGpl containing high PlsEtn concentration was prepared from oyster and ascidian, and the hydrolysates, which were rich in ly-soPlsEtn and had the same carbon chain composition as EtnGpl, were prepared from EtnGpl Fig. 1 and Tables 1 and 2 . EtnGpl administration increased the plasma levels of PlsEtn and PlsCho species, whereas administration of the hydrolysates shortened the absorption and metabolism time of plasmalogens compared to that observed with EtnGpl Figs. 2 and 3 . Moreover, hydrolysis of EtnGpl raised the AUC 0-8h of PlsEtn and PlsCho species Fig. 4 . As administration of oyster lysoEtnGpl increased the plasma levels of PlsEtn species and shortened the absorption time compared to that with oyster EtnGpl Fig. 2 , we inferred that the administered PlsEtn was hydrolyzed at the sn-2 position and re-esterized as reported in previous studies 15,16 . However, lysoPL has also been reported to affect the absorption of other lipophilic components as well as the formation of emulsion. For instance, lysoPtdCho increases the uptake of carotenoids and cholesterol in intestinal cells 24,25 ; therefore, it is also possible that lysoEtnGpl stimulates the absorption of endo-and/or exo-genous PlsEtn remaining in the digestive organs. Thus, we increased the concentration of the emulsifier in Experiment 2 relative to Experiment 1 and investigated the uptake kinetics of the mixture of EtnGpl and its hydrolysate as well as their single administrations. The uptake kinetics of their single administrations indicated tendencies that were similar to those in Experiments 1 and 2, and the administration of both Mix group indicated tendencies that were intermediate between those of EtnGpl and its hydrolysate Fig. 3 . Overall, these results strongly suggest that the administered PlsEtn is absorbed via the lysoPlsEtn form.
Administration of oyster and ascidian hydrolysates rich in lysoPlsEtn-18:0ol and EPA markedly increased the plasma levels of PlsEtn-18:0ol/20:5 from non-detectable levels before administration and also increased the levels of PlsEtn-18:0ol/20:4 and PlsCho species. Nevertheless, AA was observed in small amounts and the choline base was not detected in the samples Figs. 2 and 3 . Although DHA, EPA, AA, and 18:1 had different concentrations in the administered samples Tables 1 and 2 , PlsEtn bearing DHA, EPA, and AA showed nearly the same maximum level in plasma, whereas PlsEtn bearing 18:1 showed a slightly increased level. Previous reports show that administered PlsEtn is preferentially re-esterized to AA and partly converted to PlsCho 15,16 . On the other hand, administration of EPA-rich PlsEtn increased the serum levels of PlsEtn bearing EPA by about 40-fold compared to that before administration 26 . Our results suggest that during the absorption of PlsEtn, lysoPlsEtn was preferentially esterized to PUFA independent of the fatty acid composition in the administered samples, and that the stored range in each plasma PlsEtn species bearing PUFA may be almost at the same level.
Gradual changes in PlsEtn levels in blood help to understand the absorption and metabolism of PlsEtn, and the mechanism by which PlsEtn hydrolysis shortens absorption and metabolism duration Fig. 3 . Rate of increase in plasma PlsCho species was slower than that of PlsEtn species, and elimination of PlsCho species occurred immediately, although we could not ascertain whether the decrease was due to distribution to the organs and/or hydrolysis. While it cannot be directly compared, a previous study that investigated the maximum concentration of PlsCho species in lymph output showed a delay time after administration compared to that of the PlsEtn species due to enzymatic base conversion from PlsEtn 16 . The cardiovascular system is known to possess the highest PlsCho ratio of ChoGpl among organs 2 ; therefore, plasma PlsCho species may be easily distributed to the cardiovascular system. Additionally, administration of lysoPlsEtn decreased several plasma PtdEtn species later after administration when compared to the control group. The total level of EtnGpl is thus tightly regulated 27 . Therefore, PtdEtn biosynthesis may be suppressed due to an increase in PlsEtn. Further clarification can be accomplished by comparing the lysoPlsEtn kinetics with PlsEtn using in vitro intestinal cells and the ex vivo everted jejunal sac model.
Administration of ascidian hydrolysate rich in lyso-PlsEtn-18:0ol increased the AUC of 18:0ol-related plasmalogens until 8 h after administration compared to that of ascidian EtnGpl Fig. 4 . Continual administration of PlsEtn is reported to increase PlsEtn levels in the blood and organs of healthy rodents, and in disease models 10,12,14 . In terms of lysoPL availability, administration of lysoPtdCho with DHA has been reported to be effective in increasing the levels of PtdCho bearing DHA in rat lymph 28 . Furthermore, lysoPtdCho-DHA intake for 30 days increases DHA levels in the mouse brain 29 . Additionally, lysoPL has been recognized as a biologically active lipid mediator 30 . LysoPt-dCho-DHA shows anti-inflammatory effects in in vivo mice and in vitro macrophages 31 . Therefore, continual lyso-PlsEtn intake may effectively increase PlsEtn levels in the brain and the other organs, and thus, lysoPlsEtn may show biological activity. However, hydrolysis to lysoPlsEtn may reduce dietary functions of PlsEtn. Dietary PtdEtn and PlsEtn decrease serum cholesterol levels in rats and mice 23,32 , whereas lysoPtdCho as a PL hydrolysate accelerates cholesterol uptake in in vitro intestinal cells 24 . Additionally, it may be of interest whether lysoPlsEtn exerts the same intestinal protection as that demonstrated by PlsEtn 13 . It is thus necessary to study continual lysoPlsEtn intake in detail.
In conclusion, our results reconfirmed that administered PlsEtn was hydrolyzed at the sn-2 position and re-esterized when PlsEtn absorption occurred via lysoPlsEtn using a different approach compared to that reported in previous studies. Overall, the hydrolysis of PlsEtn may increase the PlsEtn uptake in the plasma and organs and enhance PlsEtn availability.